KumiSystems/openwrtv4
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

kernel: generic: Add kernel 4.14 support

Browse source 
This adds initial support for kernel 4.14 based on the patches for
kernel 4.9.

In the configuration I deactivated some of the new possible security
features like:
CONFIG_REFCOUNT_FULL
CONFIG_SLAB_FREELIST_HARDENED
CONFIG_SOFTLOCKUP_DETECTOR
CONFIG_WARN_ALL_UNSEEDED_RANDOM

And these overlay FS options are also deactivated:
CONFIG_OVERLAY_FS_INDEX
CONFIG_OVERLAY_FS_REDIRECT_DIR

I activated this:
CONFIG_FORTIFY_SOURCE
CONFIG_POSIX_TIMERS
CONFIG_SLAB_MERGE_DEFAULT
CONFIG_WATCHDOG_HANDLE_BOOT_ENABLED

I am not sure if I did the porting correct for the following patches:
target/linux/generic/backport-4.14/020-backport_netfilter_rtcache.patch
target/linux/generic/hack-4.14/220-gc_sections.patch
target/linux/generic/hack-4.14/321-powerpc_crtsavres_prereq.patch
target/linux/generic/pending-4.14/305-mips_module_reloc.patch
target/linux/generic/pending-4.14/611-netfilter_match_bypass_default_table.patch
target/linux/generic/pending-4.14/680-NET-skip-GRO-for-foreign-MAC-addresses.patch

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
This commit is contained in:
Hauke Mehrtens 2017-10-29 19:32:10 +01:00
parent a362df6f25
commit b3f95490b9
132 changed files with 25624 additions and 8 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
include/kernel-version.mk
View file

@ -5,10 +5,12 @@ LINUX_RELEASE?=1
LINUX_VERSION-3.18 = .71
LINUX_VERSION-4.4 = .100
LINUX_VERSION-4.9 = .67
LINUX_VERSION-4.14 = .6
LINUX_KERNEL_HASH-3.18.71 = 5abc9778ad44ce02ed6c8ab52ece8a21c6d20d21f6ed8a19287b4a38a50c1240
LINUX_KERNEL_HASH-4.4.100 = 9936cd99c4bd35f6bc6962c9acdd1fa2ac8999e07a9be6a94a03b5492f1bd14f
LINUX_KERNEL_HASH-4.9.67 = 7fbaa7dcc17877dfa0c96fb9a7d2f4ffed20ceeb13cbbeb18d77213c6cf75f7d
LINUX_KERNEL_HASH-4.14.6 = 0907678ba9ea146ddbdecd0a0b6363f56b896b5c61c9a15e809effb3ea346ccc
ifdef KERNEL_PATCHVER
LINUX_VERSION:=$(KERNEL_PATCHVER)$(strip $(LINUX_VERSION-$(KERNEL_PATCHVER)))

30
target/linux/generic/backport-4.14/010-Kbuild-don-t-hardcode-path-to-awk-in-scripts-ld-vers.patch Normal file
View file

@ -0,0 +1,30 @@
From 13b1ecc3401653a355798eb1dee10cc1608202f4 Mon Sep 17 00:00:00 2001
From: Felix Fietkau <nbd@nbd.name>
Date: Mon, 18 Jan 2016 12:27:49 +0100
Subject: [PATCH 33/34] Kbuild: don't hardcode path to awk in
scripts/ld-version.sh
On some systems /usr/bin/awk does not exist, or is broken. Find it via
$PATH instead.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
scripts/ld-version.sh | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
--- a/scripts/ld-version.sh
+++ b/scripts/ld-version.sh
@@ -1,6 +1,7 @@
-#!/usr/bin/awk -f
+#!/bin/sh
# SPDX-License-Identifier: GPL-2.0
# extract linker version number from stdin and turn into single number
+exec awk '
{
gsub(".*\\)", "");
gsub(".*version ", "");
@@ -9,3 +10,4 @@
print a[1]*100000000 + a[2]*1000000 + a[3]*10000;
exit
}
+'

23
target/linux/generic/backport-4.14/011-kbuild-export-SUBARCH.patch Normal file
View file

@ -0,0 +1,23 @@
From 173019b66dcc9d68ad9333aa744dad1e369b5aa8 Mon Sep 17 00:00:00 2001
From: Felix Fietkau <nbd@nbd.name>
Date: Sun, 9 Jul 2017 00:26:53 +0200
Subject: [PATCH 34/34] kernel: add compile fix for linux 4.9 on x86
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
Makefile | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
--- a/Makefile
+++ b/Makefile
@@ -415,8 +415,8 @@ KERNELRELEASE = $(shell cat include/conf
KERNELVERSION = $(VERSION)$(if $(PATCHLEVEL),.$(PATCHLEVEL)$(if $(SUBLEVEL),.$(SUBLEVEL)))$(EXTRAVERSION)
export VERSION PATCHLEVEL SUBLEVEL KERNELRELEASE KERNELVERSION
-export ARCH SRCARCH CONFIG_SHELL HOSTCC HOSTCFLAGS CROSS_COMPILE AS LD CC
-export CPP AR NM STRIP OBJCOPY OBJDUMP HOSTLDFLAGS HOST_LOADLIBES
+export ARCH SRCARCH SUBARCH CONFIG_SHELL HOSTCC HOSTCFLAGS CROSS_COMPILE AS LD
+export CC CPP AR NM STRIP OBJCOPY OBJDUMP HOSTLDFLAGS HOST_LOADLIBES
export MAKE AWK GENKSYMS INSTALLKERNEL PERL PYTHON UTS_MACHINE
export HOSTCXX HOSTCXXFLAGS LDFLAGS_MODULE CHECK CHECKFLAGS

558
target/linux/generic/backport-4.14/020-backport_netfilter_rtcache.patch Normal file
View file

@ -0,0 +1,558 @@
From 1bb0c3ec899827cfa4668bb63a08713a40744d21 Mon Sep 17 00:00:00 2001
From: Florian Westphal <fw@strlen.de>
Date: Sun, 9 Jul 2017 08:58:30 +0200
Subject: [PATCH] netfilter: conntrack: cache route for forwarded connections
... to avoid per-packet FIB lookup if possible.
The cached dst is re-used provided the input interface
is the same as that of the previous packet in the same direction.
If not, the cached dst is invalidated.
For ipv6 we also need to store sernum, else dst_check doesn't work,
pointed out by Eric Dumazet.
This should speed up forwarding when conntrack is already in use
anyway, especially when using reverse path filtering -- active RPF
enforces two FIB lookups for each packet.
Before the routing cache removal this didn't matter since RPF was performed
only when route cache didn't yield a result; but without route cache it
comes at higher price.
Julian Anastasov suggested to add NETDEV_UNREGISTER handler to
avoid holding on to dsts of 'frozen' conntracks.
Signed-off-by: Florian Westphal <fw@strlen.de>
---
include/net/netfilter/nf_conntrack_extend.h | 4 +
include/net/netfilter/nf_conntrack_rtcache.h | 34 +++
net/netfilter/Kconfig | 12 +
net/netfilter/Makefile | 3 +
net/netfilter/nf_conntrack_rtcache.c | 428 +++++++++++++++++++++++++++
5 files changed, 481 insertions(+)
create mode 100644 include/net/netfilter/nf_conntrack_rtcache.h
create mode 100644 net/netfilter/nf_conntrack_rtcache.c
--- a/include/net/netfilter/nf_conntrack_extend.h
+++ b/include/net/netfilter/nf_conntrack_extend.h
@@ -28,6 +28,9 @@ enum nf_ct_ext_id {
#if IS_ENABLED(CONFIG_NETFILTER_SYNPROXY)
NF_CT_EXT_SYNPROXY,
#endif
+#if IS_ENABLED(CONFIG_NF_CONNTRACK_RTCACHE)
+ NF_CT_EXT_RTCACHE,
+#endif
NF_CT_EXT_NUM,
};
@@ -40,6 +43,7 @@ enum nf_ct_ext_id {
#define NF_CT_EXT_TIMEOUT_TYPE struct nf_conn_timeout
#define NF_CT_EXT_LABELS_TYPE struct nf_conn_labels
#define NF_CT_EXT_SYNPROXY_TYPE struct nf_conn_synproxy
+#define NF_CT_EXT_RTCACHE_TYPE struct nf_conn_rtcache
/* Extensions: optional stuff which isn't permanently in struct. */
struct nf_ct_ext {
--- /dev/null
+++ b/include/net/netfilter/nf_conntrack_rtcache.h
@@ -0,0 +1,34 @@
+#include <linux/gfp.h>
+#include <net/netfilter/nf_conntrack.h>
+#include <net/netfilter/nf_conntrack_extend.h>
+
+struct dst_entry;
+
+struct nf_conn_dst_cache {
+ struct dst_entry *dst;
+ int iif;
+#if IS_ENABLED(CONFIG_NF_CONNTRACK_IPV6)
+ u32 cookie;
+#endif
+
+};
+
+struct nf_conn_rtcache {
+ struct nf_conn_dst_cache cached_dst[IP_CT_DIR_MAX];
+};
+
+static inline
+struct nf_conn_rtcache *nf_ct_rtcache_find(const struct nf_conn *ct)
+{
+#if IS_ENABLED(CONFIG_NF_CONNTRACK_RTCACHE)
+ return nf_ct_ext_find(ct, NF_CT_EXT_RTCACHE);
+#else
+ return NULL;
+#endif
+}
+
+static inline int nf_conn_rtcache_iif_get(const struct nf_conn_rtcache *rtc,
+ enum ip_conntrack_dir dir)
+{
+ return rtc->cached_dst[dir].iif;
+}
--- a/net/netfilter/Kconfig
+++ b/net/netfilter/Kconfig
@@ -118,6 +118,18 @@ config NF_CONNTRACK_EVENTS
If unsure, say `N'.
+config NF_CONNTRACK_RTCACHE
+ tristate "Cache route entries in conntrack objects"
+ depends on NETFILTER_ADVANCED
+ depends on NF_CONNTRACK
+ help
+ If this option is enabled, the connection tracking code will
+ cache routing information for each connection that is being
+ forwarded, at a cost of 32 bytes per conntrack object.
+
+ To compile it as a module, choose M here. If unsure, say N.
+ The module will be called nf_conntrack_rtcache.
+
config NF_CONNTRACK_TIMEOUT
bool 'Connection tracking timeout'
depends on NETFILTER_ADVANCED
--- a/net/netfilter/Makefile
+++ b/net/netfilter/Makefile
@@ -19,6 +19,9 @@ obj-$(CONFIG_NETFILTER_NETLINK_LOG) += n
# connection tracking
obj-$(CONFIG_NF_CONNTRACK) += nf_conntrack.o
+# optional conntrack route cache extension
+obj-$(CONFIG_NF_CONNTRACK_RTCACHE) += nf_conntrack_rtcache.o
+
obj-$(CONFIG_NF_CT_PROTO_GRE) += nf_conntrack_proto_gre.o
# netlink interface for nf_conntrack
--- /dev/null
+++ b/net/netfilter/nf_conntrack_rtcache.c
@@ -0,0 +1,428 @@
+/* route cache for netfilter.
+ *
+ * (C) 2014 Red Hat GmbH
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+
+#include <linux/types.h>
+#include <linux/netfilter.h>
+#include <linux/skbuff.h>
+#include <linux/stddef.h>
+#include <linux/kernel.h>
+#include <linux/netdevice.h>
+#include <linux/export.h>
+#include <linux/module.h>
+
+#include <net/dst.h>
+
+#include <net/netfilter/nf_conntrack.h>
+#include <net/netfilter/nf_conntrack_core.h>
+#include <net/netfilter/nf_conntrack_extend.h>
+#include <net/netfilter/nf_conntrack_rtcache.h>
+
+#if IS_ENABLED(CONFIG_NF_CONNTRACK_IPV6)
+#include <net/ip6_fib.h>
+#endif
+
+static void __nf_conn_rtcache_destroy(struct nf_conn_rtcache *rtc,
+ enum ip_conntrack_dir dir)
+{
+ struct dst_entry *dst = rtc->cached_dst[dir].dst;
+
+ dst_release(dst);
+}
+
+static void nf_conn_rtcache_destroy(struct nf_conn *ct)
+{
+ struct nf_conn_rtcache *rtc = nf_ct_rtcache_find(ct);
+
+ if (!rtc)
+ return;
+
+ __nf_conn_rtcache_destroy(rtc, IP_CT_DIR_ORIGINAL);
+ __nf_conn_rtcache_destroy(rtc, IP_CT_DIR_REPLY);
+}
+
+static void nf_ct_rtcache_ext_add(struct nf_conn *ct)
+{
+ struct nf_conn_rtcache *rtc;
+
+ rtc = nf_ct_ext_add(ct, NF_CT_EXT_RTCACHE, GFP_ATOMIC);
+ if (rtc) {
+ rtc->cached_dst[IP_CT_DIR_ORIGINAL].iif = -1;
+ rtc->cached_dst[IP_CT_DIR_ORIGINAL].dst = NULL;
+ rtc->cached_dst[IP_CT_DIR_REPLY].iif = -1;
+ rtc->cached_dst[IP_CT_DIR_REPLY].dst = NULL;
+ }
+}
+
+static struct nf_conn_rtcache *nf_ct_rtcache_find_usable(struct nf_conn *ct)
+{
+ return nf_ct_rtcache_find(ct);
+}
+
+static struct dst_entry *
+nf_conn_rtcache_dst_get(const struct nf_conn_rtcache *rtc,
+ enum ip_conntrack_dir dir)
+{
+ return rtc->cached_dst[dir].dst;
+}
+
+static u32 nf_rtcache_get_cookie(int pf, const struct dst_entry *dst)
+{
+#if IS_ENABLED(CONFIG_NF_CONNTRACK_IPV6)
+ if (pf == NFPROTO_IPV6) {
+ const struct rt6_info *rt = (const struct rt6_info *)dst;
+
+ if (rt->rt6i_node)
+ return (u32)rt->rt6i_node->fn_sernum;
+ }
+#endif
+ return 0;
+}
+
+static void nf_conn_rtcache_dst_set(int pf,
+ struct nf_conn_rtcache *rtc,
+ struct dst_entry *dst,
+ enum ip_conntrack_dir dir, int iif)
+{
+ if (rtc->cached_dst[dir].iif != iif)
+ rtc->cached_dst[dir].iif = iif;
+
+ if (rtc->cached_dst[dir].dst != dst) {
+ struct dst_entry *old;
+
+ dst_hold(dst);
+
+ old = xchg(&rtc->cached_dst[dir].dst, dst);
+ dst_release(old);
+
+#if IS_ENABLED(CONFIG_NF_CONNTRACK_IPV6)
+ if (pf == NFPROTO_IPV6)
+ rtc->cached_dst[dir].cookie =
+ nf_rtcache_get_cookie(pf, dst);
+#endif
+ }
+}
+
+static void nf_conn_rtcache_dst_obsolete(struct nf_conn_rtcache *rtc,
+ enum ip_conntrack_dir dir)
+{
+ struct dst_entry *old;
+
+ pr_debug("Invalidate iif %d for dir %d on cache %p\n",
+ rtc->cached_dst[dir].iif, dir, rtc);
+
+ old = xchg(&rtc->cached_dst[dir].dst, NULL);
+ dst_release(old);
+ rtc->cached_dst[dir].iif = -1;
+}
+
+static unsigned int nf_rtcache_in(u_int8_t pf,
+ struct sk_buff *skb,
+ const struct nf_hook_state *state)
+{
+ struct nf_conn_rtcache *rtc;
+ enum ip_conntrack_info ctinfo;
+ enum ip_conntrack_dir dir;
+ struct dst_entry *dst;
+ struct nf_conn *ct;
+ int iif;
+ u32 cookie;
+
+ if (skb_dst(skb) || skb->sk)
+ return NF_ACCEPT;
+
+ ct = nf_ct_get(skb, &ctinfo);
+ if (!ct)
+ return NF_ACCEPT;
+
+ rtc = nf_ct_rtcache_find_usable(ct);
+ if (!rtc)
+ return NF_ACCEPT;
+
+ /* if iif changes, don't use cache and let ip stack
+ * do route lookup.
+ *
+ * If rp_filter is enabled it might toss skb, so
+ * we don't want to avoid these checks.
+ */
+ dir = CTINFO2DIR(ctinfo);
+ iif = nf_conn_rtcache_iif_get(rtc, dir);
+ if (state->in->ifindex != iif) {
+ pr_debug("ct %p, iif %d, cached iif %d, skip cached entry\n",
+ ct, iif, state->in->ifindex);
+ return NF_ACCEPT;
+ }
+ dst = nf_conn_rtcache_dst_get(rtc, dir);
+ if (dst == NULL)
+ return NF_ACCEPT;
+
+ cookie = nf_rtcache_get_cookie(pf, dst);
+
+ dst = dst_check(dst, cookie);
+ pr_debug("obtained dst %p for skb %p, cookie %d\n", dst, skb, cookie);
+ if (likely(dst))
+ skb_dst_set_noref(skb, dst);
+ else
+ nf_conn_rtcache_dst_obsolete(rtc, dir);
+
+ return NF_ACCEPT;
+}
+
+static unsigned int nf_rtcache_forward(u_int8_t pf,
+ struct sk_buff *skb,
+ const struct nf_hook_state *state)
+{
+ struct nf_conn_rtcache *rtc;
+ enum ip_conntrack_info ctinfo;
+ enum ip_conntrack_dir dir;
+ struct nf_conn *ct;
+ struct dst_entry *dst = skb_dst(skb);
+ int iif;
+
+ ct = nf_ct_get(skb, &ctinfo);
+ if (!ct)
+ return NF_ACCEPT;
+
+ if (dst && dst_xfrm(dst))
+ return NF_ACCEPT;
+
+ if (!nf_ct_is_confirmed(ct)) {
+ if (WARN_ON(nf_ct_rtcache_find(ct)))
+ return NF_ACCEPT;
+ nf_ct_rtcache_ext_add(ct);
+ return NF_ACCEPT;
+ }
+
+ rtc = nf_ct_rtcache_find_usable(ct);
+ if (!rtc)
+ return NF_ACCEPT;
+
+ dir = CTINFO2DIR(ctinfo);
+ iif = nf_conn_rtcache_iif_get(rtc, dir);
+ pr_debug("ct %p, skb %p, dir %d, iif %d, cached iif %d\n",
+ ct, skb, dir, iif, state->in->ifindex);
+ if (likely(state->in->ifindex == iif))
+ return NF_ACCEPT;
+
+ nf_conn_rtcache_dst_set(pf, rtc, skb_dst(skb), dir, state->in->ifindex);
+ return NF_ACCEPT;
+}
+
+static unsigned int nf_rtcache_in4(void *priv,
+ struct sk_buff *skb,
+ const struct nf_hook_state *state)
+{
+ return nf_rtcache_in(NFPROTO_IPV4, skb, state);
+}
+
+static unsigned int nf_rtcache_forward4(void *priv,
+ struct sk_buff *skb,
+ const struct nf_hook_state *state)
+{
+ return nf_rtcache_forward(NFPROTO_IPV4, skb, state);
+}
+
+#if IS_ENABLED(CONFIG_NF_CONNTRACK_IPV6)
+static unsigned int nf_rtcache_in6(void *priv,
+ struct sk_buff *skb,
+ const struct nf_hook_state *state)
+{
+ return nf_rtcache_in(NFPROTO_IPV6, skb, state);
+}
+
+static unsigned int nf_rtcache_forward6(void *priv,
+ struct sk_buff *skb,
+ const struct nf_hook_state *state)
+{
+ return nf_rtcache_forward(NFPROTO_IPV6, skb, state);
+}
+#endif
+
+static int nf_rtcache_dst_remove(struct nf_conn *ct, void *data)
+{
+ struct nf_conn_rtcache *rtc = nf_ct_rtcache_find(ct);
+ struct net_device *dev = data;
+
+ if (!rtc)
+ return 0;
+
+ if (dev->ifindex == rtc->cached_dst[IP_CT_DIR_ORIGINAL].iif ||
+ dev->ifindex == rtc->cached_dst[IP_CT_DIR_REPLY].iif) {
+ nf_conn_rtcache_dst_obsolete(rtc, IP_CT_DIR_ORIGINAL);
+ nf_conn_rtcache_dst_obsolete(rtc, IP_CT_DIR_REPLY);
+ }
+
+ return 0;
+}
+
+static int nf_rtcache_netdev_event(struct notifier_block *this,
+ unsigned long event, void *ptr)
+{
+ struct net_device *dev = netdev_notifier_info_to_dev(ptr);
+ struct net *net = dev_net(dev);
+
+ if (event == NETDEV_DOWN)
+ nf_ct_iterate_cleanup_net(net, nf_rtcache_dst_remove, dev, 0, 0);
+
+ return NOTIFY_DONE;
+}
+
+static struct notifier_block nf_rtcache_notifier = {
+ .notifier_call = nf_rtcache_netdev_event,
+};
+
+static struct nf_hook_ops rtcache_ops[] = {
+ {
+ .hook = nf_rtcache_in4,
+ .pf = NFPROTO_IPV4,
+ .hooknum = NF_INET_PRE_ROUTING,
+ .priority = NF_IP_PRI_LAST,
+ },
+ {
+ .hook = nf_rtcache_forward4,
+ .pf = NFPROTO_IPV4,
+ .hooknum = NF_INET_FORWARD,
+ .priority = NF_IP_PRI_LAST,
+ },
+#if IS_ENABLED(CONFIG_NF_CONNTRACK_IPV6)
+ {
+ .hook = nf_rtcache_in6,
+ .pf = NFPROTO_IPV6,
+ .hooknum = NF_INET_PRE_ROUTING,
+ .priority = NF_IP_PRI_LAST,
+ },
+ {
+ .hook = nf_rtcache_forward6,
+ .pf = NFPROTO_IPV6,
+ .hooknum = NF_INET_FORWARD,
+ .priority = NF_IP_PRI_LAST,
+ },
+#endif
+};
+
+static struct nf_ct_ext_type rtcache_extend __read_mostly = {
+ .len = sizeof(struct nf_conn_rtcache),
+ .align = __alignof__(struct nf_conn_rtcache),
+ .id = NF_CT_EXT_RTCACHE,
+ .destroy = nf_conn_rtcache_destroy,
+};
+
+static void __net_exit rtcache_net_exit(struct net *net)
+{
+ /* remove hooks so no new connections get rtcache extension */
+ nf_unregister_net_hooks(net, rtcache_ops, ARRAY_SIZE(rtcache_ops));
+}
+
+static struct pernet_operations rtcache_ops_net_ops = {
+ .exit = rtcache_net_exit,
+};
+
+static int __init nf_conntrack_rtcache_init(void)
+{
+ int ret = nf_ct_extend_register(&rtcache_extend);
+
+ if (ret < 0) {
+ pr_err("nf_conntrack_rtcache: Unable to register extension\n");
+ return ret;
+ }
+
+ ret = register_pernet_subsys(&rtcache_ops_net_ops);
+ if (ret) {
+ nf_ct_extend_unregister(&rtcache_extend);
+ return ret;
+ }
+
+ ret = nf_register_net_hooks(&init_net, rtcache_ops,
+ ARRAY_SIZE(rtcache_ops));
+ if (ret < 0) {
+ nf_ct_extend_unregister(&rtcache_extend);
+ unregister_pernet_subsys(&rtcache_ops_net_ops);
+ return ret;
+ }
+
+ ret = register_netdevice_notifier(&nf_rtcache_notifier);
+ if (ret) {
+ nf_unregister_net_hooks(&init_net, rtcache_ops,
+ ARRAY_SIZE(rtcache_ops));
+ nf_ct_extend_unregister(&rtcache_extend);
+ unregister_pernet_subsys(&rtcache_ops_net_ops);
+ }
+
+ return ret;
+}
+
+static int nf_rtcache_ext_remove(struct nf_conn *ct, void *data)
+{
+ struct nf_conn_rtcache *rtc = nf_ct_rtcache_find(ct);
+
+ return rtc != NULL;
+}
+
+static bool __exit nf_conntrack_rtcache_wait_for_dying(struct net *net)
+{
+ bool wait = false;
+ int cpu;
+
+ for_each_possible_cpu(cpu) {
+ struct nf_conntrack_tuple_hash *h;
+ struct hlist_nulls_node *n;
+ struct nf_conn *ct;
+ struct ct_pcpu *pcpu = per_cpu_ptr(net->ct.pcpu_lists, cpu);
+
+ rcu_read_lock();
+ spin_lock_bh(&pcpu->lock);
+
+ hlist_nulls_for_each_entry(h, n, &pcpu->dying, hnnode) {
+ ct = nf_ct_tuplehash_to_ctrack(h);
+ if (nf_ct_rtcache_find(ct) != NULL) {
+ wait = true;
+ break;
+ }
+ }
+ spin_unlock_bh(&pcpu->lock);
+ rcu_read_unlock();
+ }
+
+ return wait;
+}
+
+static void __exit nf_conntrack_rtcache_fini(void)
+{
+ struct net *net;
+ int count = 0;
+
+ synchronize_net();
+
+ unregister_netdevice_notifier(&nf_rtcache_notifier);
+
+ rtnl_lock();
+
+ /* zap all conntracks with rtcache extension */
+ for_each_net(net)
+ nf_ct_iterate_cleanup_net(net, nf_rtcache_ext_remove, NULL, 0, 0);
+
+ for_each_net(net) {
+ /* .. and make sure they're gone from dying list, too */
+ while (nf_conntrack_rtcache_wait_for_dying(net)) {
+ msleep(200);
+ WARN_ONCE(++count > 25, "Waiting for all rtcache conntracks to go away\n");
+ }
+ }
+
+ rtnl_unlock();
+ synchronize_net();
+ nf_ct_extend_unregister(&rtcache_extend);
+}
+module_init(nf_conntrack_rtcache_init);
+module_exit(nf_conntrack_rtcache_fini);
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Florian Westphal <fw@strlen.de>");
+MODULE_DESCRIPTION("Conntrack route cache extension");

85
target/linux/generic/backport-4.14/025-tcp-allow-drivers-to-tweak-TSQ-logic.patch Normal file
View file

@ -0,0 +1,85 @@
From: Eric Dumazet <edumazet@google.com>
Date: Sat, 11 Nov 2017 15:54:12 -0800
Subject: [PATCH] tcp: allow drivers to tweak TSQ logic
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
I had many reports that TSQ logic breaks wifi aggregation.
Current logic is to allow up to 1 ms of bytes to be queued into qdisc
and drivers queues.
But Wifi aggregation needs a bigger budget to allow bigger rates to
be discovered by various TCP Congestion Controls algorithms.
This patch adds an extra socket field, allowing wifi drivers to select
another log scale to derive TCP Small Queue credit from current pacing
rate.
Initial value is 10, meaning that this patch does not change current
behavior.
We expect wifi drivers to set this field to smaller values (tests have
been done with values from 6 to 9)
They would have to use following template :
if (skb->sk && skb->sk->sk_pacing_shift != MY_PACING_SHIFT)
skb->sk->sk_pacing_shift = MY_PACING_SHIFT;
Ref: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1670041
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Johannes Berg <johannes.berg@intel.com>
Cc: Toke Høiland-Jørgensen <toke@toke.dk>
Cc: Kir Kolyshkin <kir@openvz.org>
---
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -267,6 +267,7 @@ struct sock_common {
* @sk_gso_type: GSO type (e.g. %SKB_GSO_TCPV4)
* @sk_gso_max_size: Maximum GSO segment size to build
* @sk_gso_max_segs: Maximum number of GSO segments
+ * @sk_pacing_shift: scaling factor for TCP Small Queues
* @sk_lingertime: %SO_LINGER l_linger setting
* @sk_backlog: always used with the per-socket spinlock held
* @sk_callback_lock: used with the callbacks in the end of this struct
@@ -448,6 +449,8 @@ struct sock {
kmemcheck_bitfield_end(flags);
u16 sk_gso_max_segs;
+#define sk_pacing_shift sk_pacing_shift /* for backport checks */
+ u8 sk_pacing_shift;
unsigned long sk_lingertime;
struct proto *sk_prot_creator;
rwlock_t sk_callback_lock;
--- a/net/core/sock.c
+++ b/net/core/sock.c
@@ -2744,6 +2744,7 @@ void sock_init_data(struct socket *sock,
sk->sk_max_pacing_rate = ~0U;
sk->sk_pacing_rate = ~0U;
+ sk->sk_pacing_shift = 10;
sk->sk_incoming_cpu = -1;
/*
* Before updating sk_refcnt, we must commit prior changes to memory
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -1671,7 +1671,7 @@ u32 tcp_tso_autosize(const struct sock *
{
u32 bytes, segs;
- bytes = min(sk->sk_pacing_rate >> 10,
+ bytes = min(sk->sk_pacing_rate >> sk->sk_pacing_shift,
sk->sk_gso_max_size - 1 - MAX_TCP_HEADER);
/* Goal is to send at least one packet per ms,
@@ -2145,7 +2145,7 @@ static bool tcp_small_queue_check(struct
{
unsigned int limit;
- limit = max(2 * skb->truesize, sk->sk_pacing_rate >> 10);
+ limit = max(2 * skb->truesize, sk->sk_pacing_rate >> sk->sk_pacing_shift);
limit = min_t(u32, limit, sysctl_tcp_limit_output_bytes);
limit <<= factor;

50
target/linux/generic/backport-4.14/071-v4.15-0001-net-bgmac-enable-master-mode-for-BCM54210E-and-B5021.patch Normal file
View file

@ -0,0 +1,50 @@
From 12acd136913ccdf394eeb2bc8686ff5505368119 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Rafa=C5=82=20Mi=C5=82ecki?= <rafal@milecki.pl>
Date: Thu, 12 Oct 2017 10:21:26 +0200
Subject: [PATCH] net: bgmac: enable master mode for BCM54210E and B50212E PHYs
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
There are 4 very similar PHYs:
0x600d84a1: BCM54210E (rev B0)
0x600d84a2: BCM54210E (rev B1)
0x600d84a5: B50212E (rev B0)
0x600d84a6: B50212E (rev B1)
that need setting master mode manually. It's because they run in slave
mode by default with Automatic Slave/Master configuration disabled which
can lead to unreliable connection with massive ping loss.
So far it was reported for a board with BCM47189 SoC and B50212E B1 PHY
connected to the bgmac supported ethernet device. Telling PHY driver to
setup PHY properly solves this issue.
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
drivers/net/ethernet/broadcom/bgmac-bcma.c | 8 +++++++-
1 file changed, 7 insertions(+), 1 deletion(-)
--- a/drivers/net/ethernet/broadcom/bgmac-bcma.c
+++ b/drivers/net/ethernet/broadcom/bgmac-bcma.c
@@ -184,13 +184,19 @@ static int bgmac_probe(struct bcma_devic
if (!bgmac_is_bcm4707_family(core) &&
!(ci->id == BCMA_CHIP_ID_BCM53573 && core->core_unit == 1)) {
+ struct phy_device *phydev;
+
mii_bus = bcma_mdio_mii_register(bgmac);
if (IS_ERR(mii_bus)) {
err = PTR_ERR(mii_bus);
goto err;
}
-
bgmac->mii_bus = mii_bus;
+
+ phydev = mdiobus_get_phy(bgmac->mii_bus, bgmac->phyaddr);
+ if (ci->id == BCMA_CHIP_ID_BCM53573 && phydev &&
+ (phydev->drv->phy_id & phydev->drv->phy_id_mask) == PHY_ID_BCM54210E)
+ phydev->dev_flags |= PHY_BRCM_EN_MASTER_MODE;
}
if (core->bus->hosttype == BCMA_HOSTTYPE_PCI) {

54
target/linux/generic/backport-4.14/076-v4.15-0001-net-phy-broadcom-support-new-device-flag-for-setting.patch Normal file
View file

@ -0,0 +1,54 @@
From 2355a6546a053b1c16ebefd6ce1f0cccc00e1da5 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Rafa=C5=82=20Mi=C5=82ecki?= <rafal@milecki.pl>
Date: Thu, 12 Oct 2017 10:21:25 +0200
Subject: [PATCH] net: phy: broadcom: support new device flag for setting
master mode
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Some of Broadcom's PHYs run by default in slave mode with Automatic
Slave/Master configuration disabled. It stops them from working properly
with some devices.
So far it has been verified for BCM54210E and BCM50212E which don't
work well with Intel's I217-LM and I218-LM:
http://ark.intel.com/products/60019/Intel-Ethernet-Connection-I217-LM
http://ark.intel.com/products/71307/Intel-Ethernet-Connection-I218-LM
I was told there is massive ping loss.
This commit adds support for a new flag which can be set by an ethernet
driver to fixup PHY setup.
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
drivers/net/phy/broadcom.c | 6 ++++++
include/linux/brcmphy.h | 1 +
2 files changed, 7 insertions(+)
--- a/drivers/net/phy/broadcom.c
+++ b/drivers/net/phy/broadcom.c
@@ -43,6 +43,12 @@ static int bcm54210e_config_init(struct
val &= ~BCM54810_SHD_CLK_CTL_GTXCLK_EN;
bcm_phy_write_shadow(phydev, BCM54810_SHD_CLK_CTL, val);
+ if (phydev->dev_flags & PHY_BRCM_EN_MASTER_MODE) {
+ val = phy_read(phydev, MII_CTRL1000);
+ val |= CTL1000_AS_MASTER | CTL1000_ENABLE_MASTER;
+ phy_write(phydev, MII_CTRL1000, val);
+ }
+
return 0;
}
--- a/include/linux/brcmphy.h
+++ b/include/linux/brcmphy.h
@@ -64,6 +64,7 @@
#define PHY_BRCM_EXT_IBND_TX_ENABLE 0x00002000
#define PHY_BRCM_CLEAR_RGMII_MODE 0x00004000
#define PHY_BRCM_DIS_TXCRXC_NOENRGY 0x00008000
+#define PHY_BRCM_EN_MASTER_MODE 0x00010000
/* Broadcom BCM7xxx specific workarounds */
#define PHY_BRCM_7XXX_REV(x) (((x) >> 8) & 0xff)

5591
target/linux/generic/config-4.14 Normal file
View file

File diff suppressed because it is too large Load diff

1
target/linux/generic/files/drivers/net/phy/b53/b53_common.c
View file

@ -24,6 +24,7 @@
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/switch.h>
#include <linux/phy.h>
#include <linux/of.h>
#include <linux/of_net.h>
#include <linux/platform_data/b53.h>

3
target/linux/generic/files/drivers/net/phy/rtl8306.c
View file

@ -25,6 +25,7 @@
#include <linux/switch.h>
#include <linux/delay.h>
#include <linux/phy.h>
#include <linux/version.h>
//#define DEBUG 1
@ -1032,7 +1033,9 @@ rtl8306_read_status(struct phy_device *pdev)
static struct phy_driver rtl8306_driver = {
.name = "Realtek RTL8306S",
#if (LINUX_VERSION_CODE < KERNEL_VERSION(4,13,0))
.flags = PHY_HAS_MAGICANEG,
#endif
.phy_id = RTL8306_MAGIC,
.phy_id_mask = 0xffffffff,
.features = PHY_BASIC_FEATURES,

34
target/linux/generic/files/drivers/net/phy/swconfig.c
View file

@ -269,13 +269,7 @@ static void swconfig_defaults_init(struct switch_dev *dev)
}
static struct genl_family switch_fam = {
.id = GENL_ID_GENERATE,
.name = "switch",
.hdrsize = 0,
.version = 1,
.maxattr = SWITCH_ATTR_MAX,
};
static struct genl_family switch_fam;
static const struct nla_policy switch_policy[SWITCH_ATTR_MAX+1] = {
[SWITCH_ATTR_ID] = { .type = NLA_U32 },
@ -597,8 +591,13 @@ swconfig_parse_ports(struct sk_buff *msg, struct nlattr *head,
port = &val->value.ports[val->len];
#if LINUX_VERSION_CODE < KERNEL_VERSION(4,12,0)
if (nla_parse_nested(tb, SWITCH_PORT_ATTR_MAX, nla,
port_policy))
#else
if (nla_parse_nested(tb, SWITCH_PORT_ATTR_MAX, nla,
port_policy, NULL))
#endif
return -EINVAL;
if (!tb[SWITCH_PORT_ID])
@ -619,7 +618,11 @@ swconfig_parse_link(struct sk_buff *msg, struct nlattr *nla,
{
struct nlattr *tb[SWITCH_LINK_ATTR_MAX + 1];
#if LINUX_VERSION_CODE < KERNEL_VERSION(4,12,0)
if (nla_parse_nested(tb, SWITCH_LINK_ATTR_MAX, nla, link_policy))
#else
if (nla_parse_nested(tb, SWITCH_LINK_ATTR_MAX, nla, link_policy, NULL))
#endif
return -EINVAL;
link->duplex = !!tb[SWITCH_LINK_FLAG_DUPLEX];
@ -1051,6 +1054,19 @@ static struct genl_ops swconfig_ops[] = {
}
};
static struct genl_family switch_fam = {
#if LINUX_VERSION_CODE < KERNEL_VERSION(4,10,0)
.id = GENL_ID_GENERATE,
#endif
.name = "switch",
.hdrsize = 0,
.version = 1,
.maxattr = SWITCH_ATTR_MAX,
.module = THIS_MODULE,
.ops = swconfig_ops,
.n_ops = ARRAY_SIZE(swconfig_ops),
};
#ifdef CONFIG_OF
void
of_switch_load_portmap(struct switch_dev *dev)
@ -1223,7 +1239,11 @@ swconfig_init(void)
{
INIT_LIST_HEAD(&swdevs);
#if LINUX_VERSION_CODE < KERNEL_VERSION(4,10,0)
return genl_register_family_with_ops(&switch_fam, swconfig_ops);
#else
return genl_register_family(&switch_fam);
#endif
}
static void __exit

24
target/linux/generic/hack-4.14/202-reduce_module_size.patch Normal file
View file

@ -0,0 +1,24 @@
From fd66884da2f96d2a7ea73f58b1b86251b959a913 Mon Sep 17 00:00:00 2001
From: Felix Fietkau <nbd@nbd.name>
Date: Fri, 7 Jul 2017 16:56:19 +0200
Subject: kernel: strip unnecessary symbol table information from kernel modules
reduces default squashfs size on ar71xx by about 4k
lede-commit: 058d331a39077f159ca8922f1f422a1346d6aa67
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
Makefile | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
--- a/Makefile
+++ b/Makefile
@@ -408,7 +408,7 @@ KBUILD_CFLAGS_KERNEL :=
KBUILD_AFLAGS := -D__ASSEMBLY__ $(call cc-option,-fno-PIE)
KBUILD_AFLAGS_MODULE := -DMODULE
KBUILD_CFLAGS_MODULE := -DMODULE
-KBUILD_LDFLAGS_MODULE := -T $(srctree)/scripts/module-common.lds
+KBUILD_LDFLAGS_MODULE = -T $(srctree)/scripts/module-common.lds $(if $(CONFIG_PROFILING),,-s)
# Read KERNELRELEASE from include/config/kernel.release (if it exists)
KERNELRELEASE = $(shell cat include/config/kernel.release 2> /dev/null)

205
target/linux/generic/hack-4.14/204-module_strip.patch Normal file
View file

@ -0,0 +1,205 @@
From a779a482fb9b9f8fcdf8b2519c789b4b9bb5dd05 Mon Sep 17 00:00:00 2001
From: Felix Fietkau <nbd@nbd.name>
Date: Fri, 7 Jul 2017 16:56:48 +0200
Subject: build: add a hack for removing non-essential module info
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
include/linux/module.h | 13 ++++++++-----
include/linux/moduleparam.h | 15 ++++++++++++---
init/Kconfig | 7 +++++++
kernel/module.c | 5 ++++-
scripts/mod/modpost.c | 12 ++++++++++++
5 files changed, 43 insertions(+), 9 deletions(-)
--- a/include/linux/module.h
+++ b/include/linux/module.h
@@ -158,6 +158,7 @@ extern void cleanup_module(void);
/* Generic info of form tag = "info" */
#define MODULE_INFO(tag, info) __MODULE_INFO(tag, tag, info)
+#define MODULE_INFO_STRIP(tag, info) __MODULE_INFO_STRIP(tag, tag, info)
/* For userspace: you can also call me... */
#define MODULE_ALIAS(_alias) MODULE_INFO(alias, _alias)
@@ -201,12 +202,12 @@ extern void cleanup_module(void);
* Author(s), use "Name <email>" or just "Name", for multiple
* authors use multiple MODULE_AUTHOR() statements/lines.
*/
-#define MODULE_AUTHOR(_author) MODULE_INFO(author, _author)
+#define MODULE_AUTHOR(_author) MODULE_INFO_STRIP(author, _author)
/* What your module does. */
-#define MODULE_DESCRIPTION(_description) MODULE_INFO(description, _description)
+#define MODULE_DESCRIPTION(_description) MODULE_INFO_STRIP(description, _description)
-#ifdef MODULE
+#if defined(MODULE) && !defined(CONFIG_MODULE_STRIPPED)
/* Creates an alias so file2alias.c can find device table. */
#define MODULE_DEVICE_TABLE(type, name) \
extern typeof(name) __mod_##type##__##name##_device_table \
@@ -233,7 +234,9 @@ extern typeof(name) __mod_##type##__##na
*/
#if defined(MODULE) || !defined(CONFIG_SYSFS)
-#define MODULE_VERSION(_version) MODULE_INFO(version, _version)
+#define MODULE_VERSION(_version) MODULE_INFO_STRIP(version, _version)
+#elif defined(CONFIG_MODULE_STRIPPED)
+#define MODULE_VERSION(_version) __MODULE_INFO_DISABLED(version)
#else
#define MODULE_VERSION(_version) \
static struct module_version_attribute ___modver_attr = { \
@@ -255,7 +258,7 @@ extern typeof(name) __mod_##type##__##na
/* Optional firmware file (or files) needed by the module
* format is simply firmware file name. Multiple firmware
* files require multiple MODULE_FIRMWARE() specifiers */
-#define MODULE_FIRMWARE(_firmware) MODULE_INFO(firmware, _firmware)
+#define MODULE_FIRMWARE(_firmware) MODULE_INFO_STRIP(firmware, _firmware)
struct notifier_block;
--- a/include/linux/moduleparam.h
+++ b/include/linux/moduleparam.h
@@ -17,6 +17,16 @@
/* Chosen so that structs with an unsigned long line up. */
#define MAX_PARAM_PREFIX_LEN (64 - sizeof(unsigned long))
+/* This struct is here for syntactic coherency, it is not used */
+#define __MODULE_INFO_DISABLED(name) \
+ struct __UNIQUE_ID(name) {}
+
+#ifdef CONFIG_MODULE_STRIPPED
+#define __MODULE_INFO_STRIP(tag, name, info) __MODULE_INFO_DISABLED(name)
+#else
+#define __MODULE_INFO_STRIP(tag, name, info) __MODULE_INFO(tag, name, info)
+#endif
+
#ifdef MODULE
#define __MODULE_INFO(tag, name, info) \
static const char __UNIQUE_ID(name)[] \
@@ -24,8 +34,7 @@ static const char __UNIQUE_ID(name)[]
= __stringify(tag) "=" info
#else /* !MODULE */
/* This struct is here for syntactic coherency, it is not used */
-#define __MODULE_INFO(tag, name, info) \
- struct __UNIQUE_ID(name) {}
+#define __MODULE_INFO(tag, name, info) __MODULE_INFO_DISABLED(name)
#endif
#define __MODULE_PARM_TYPE(name, _type) \
__MODULE_INFO(parmtype, name##type, #name ":" _type)
@@ -33,7 +42,7 @@ static const char __UNIQUE_ID(name)[]
/* One for each parameter, describing how to use it. Some files do
multiple of these per line, so can't just use MODULE_INFO. */
#define MODULE_PARM_DESC(_parm, desc) \
- __MODULE_INFO(parm, _parm, #_parm ":" desc)
+ __MODULE_INFO_STRIP(parm, _parm, #_parm ":" desc)
struct kernel_param;
--- a/init/Kconfig
+++ b/init/Kconfig
@@ -1896,6 +1896,13 @@ config TRIM_UNUSED_KSYMS
If unsure, or if you need to build out-of-tree modules, say N.
+config MODULE_STRIPPED
+ bool "Reduce module size"
+ depends on MODULES
+ help
+ Remove module parameter descriptions, author info, version, aliases,
+ device tables, etc.
+
endif # MODULES
config MODULES_TREE_LOOKUP
--- a/kernel/module.c
+++ b/kernel/module.c
@@ -2997,9 +2997,11 @@ static struct module *setup_load_info(st
static int check_modinfo(struct module *mod, struct load_info *info, int flags)
{
- const char *modmagic = get_modinfo(info, "vermagic");
int err;
+#ifndef CONFIG_MODULE_STRIPPED
+ const char *modmagic = get_modinfo(info, "vermagic");
+
if (flags & MODULE_INIT_IGNORE_VERMAGIC)
modmagic = NULL;
@@ -3020,6 +3022,7 @@ static int check_modinfo(struct module *
mod->name);
add_taint_module(mod, TAINT_OOT_MODULE, LOCKDEP_STILL_OK);
}
+#endif
if (get_modinfo(info, "staging")) {
add_taint_module(mod, TAINT_CRAP, LOCKDEP_STILL_OK);
--- a/scripts/mod/modpost.c
+++ b/scripts/mod/modpost.c
@@ -1982,7 +1982,9 @@ static void read_symbols(char *modname)
symname = remove_dot(info.strtab + sym->st_name);
handle_modversions(mod, &info, sym, symname);
+#ifndef CONFIG_MODULE_STRIPPED
handle_moddevtable(mod, &info, sym, symname);
+#endif
}
if (!is_vmlinux(modname) ||
(is_vmlinux(modname) && vmlinux_section_warnings))
@@ -2143,8 +2145,10 @@ static void add_header(struct buffer *b,
buf_printf(b, "#include <linux/vermagic.h>\n");
buf_printf(b, "#include <linux/compiler.h>\n");
buf_printf(b, "\n");
+#ifndef CONFIG_MODULE_STRIPPED
buf_printf(b, "MODULE_INFO(vermagic, VERMAGIC_STRING);\n");
buf_printf(b, "MODULE_INFO(name, KBUILD_MODNAME);\n");
+#endif
buf_printf(b, "\n");
buf_printf(b, "__visible struct module __this_module\n");
buf_printf(b, "__attribute__((section(\".gnu.linkonce.this_module\"))) = {\n");
@@ -2161,16 +2165,20 @@ static void add_header(struct buffer *b,
static void add_intree_flag(struct buffer *b, int is_intree)
{
+#ifndef CONFIG_MODULE_STRIPPED
if (is_intree)
buf_printf(b, "\nMODULE_INFO(intree, \"Y\");\n");
+#endif
}
static void add_staging_flag(struct buffer *b, const char *name)
{
+#ifndef CONFIG_MODULE_STRIPPED
static const char *staging_dir = "drivers/staging";
if (strncmp(staging_dir, name, strlen(staging_dir)) == 0)
buf_printf(b, "\nMODULE_INFO(staging, \"Y\");\n");
+#endif
}
/**
@@ -2269,11 +2277,13 @@ static void add_depends(struct buffer *b
static void add_srcversion(struct buffer *b, struct module *mod)
{
+#ifndef CONFIG_MODULE_STRIPPED
if (mod->srcversion[0]) {
buf_printf(b, "\n");
buf_printf(b, "MODULE_INFO(srcversion, \"%s\");\n",
mod->srcversion);
}
+#endif
}
static void write_if_changed(struct buffer *b, const char *fname)
@@ -2509,7 +2519,9 @@ int main(int argc, char **argv)
add_staging_flag(&buf, mod->name);
err |= add_versions(&buf, mod);
add_depends(&buf, mod, modules);
+#ifndef CONFIG_MODULE_STRIPPED
add_moddevtable(&buf, mod);
+#endif
add_srcversion(&buf, mod);
sprintf(fname, "%s.mod.c", mod->name);

44
target/linux/generic/hack-4.14/207-disable-modorder.patch Normal file
View file

@ -0,0 +1,44 @@
From c9ef4ab0f54356ee9f91d9676ea0ec123840ddc7 Mon Sep 17 00:00:00 2001
From: Felix Fietkau <nbd@nbd.name>
Date: Fri, 7 Jul 2017 16:57:33 +0200
Subject: kernel: do not build modules.order
It is not needed for anything on the system and skipping this saves some
build time, especially in cases where there is nothing to do.
lede-commit: afc1675833a7bf5df094f59f7250369520646d04
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
Makefile | 2 --
scripts/Makefile.build | 2 +-
2 files changed, 1 insertion(+), 3 deletions(-)
--- a/Makefile
+++ b/Makefile
@@ -1208,7 +1208,6 @@ all: modules
PHONY += modules
modules: $(vmlinux-dirs) $(if $(KBUILD_BUILTIN),vmlinux) modules.builtin
- $(Q)$(AWK) '!x[$$0]++' $(vmlinux-dirs:%=$(objtree)/%/modules.order) > $(objtree)/modules.order
@$(kecho) ' Building modules, stage 2.';
$(Q)$(MAKE) -f $(srctree)/scripts/Makefile.modpost
@@ -1237,7 +1236,6 @@ _modinst_:
rm -f $(MODLIB)/build ; \
ln -s $(CURDIR) $(MODLIB)/build ; \
fi
- @cp -f $(objtree)/modules.order $(MODLIB)/
@cp -f $(objtree)/modules.builtin $(MODLIB)/
$(Q)$(MAKE) -f $(srctree)/scripts/Makefile.modinst
--- a/scripts/Makefile.build
+++ b/scripts/Makefile.build
@@ -94,7 +94,7 @@ modorder-target := $(obj)/modules.order
# We keep a list of all modules in $(MODVERDIR)
__build: $(if $(KBUILD_BUILTIN),$(builtin-target) $(lib-target) $(extra-y)) \
- $(if $(KBUILD_MODULES),$(obj-m) $(modorder-target)) \
+ $(if $(KBUILD_MODULES),$(obj-m)) \
$(subdir-ym) $(always)
@:

3065
target/linux/generic/hack-4.14/210-darwin_scripts_include.patch Normal file
View file

File diff suppressed because it is too large Load diff

40
target/linux/generic/hack-4.14/211-host_tools_portability.patch Normal file
View file

@ -0,0 +1,40 @@
From 7f698012384ccb1ed10cc758acfd085096fdb307 Mon Sep 17 00:00:00 2001
From: Felix Fietkau <nbd@nbd.name>
Date: Fri, 7 Jul 2017 17:02:03 +0200
Subject: kernel: fix linux 4.9 host tools portability issues
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
tools/build/Build.include | 2 +-
tools/perf/pmu-events/jevents.c | 1 +
tools/perf/pmu-events/json.c | 1 -
3 files changed, 2 insertions(+), 2 deletions(-)
--- a/tools/build/Build.include
+++ b/tools/build/Build.include
@@ -97,4 +97,4 @@ cxx_flags = -Wp,-MD,$(depfile) -Wp,-MT,$
###
## HOSTCC C flags
-host_c_flags = -Wp,-MD,$(depfile) -Wp,-MT,$@ $(CHOSTFLAGS) -D"BUILD_STR(s)=\#s" $(CHOSTFLAGS_$(basetarget).o) $(CHOSTFLAGS_$(obj))
+host_c_flags = -MD -MF $(depfile) -Wp -MT $@ $(CHOSTFLAGS) -D"BUILD_STR(s)=\#s" $(CHOSTFLAGS_$(basetarget).o) $(CHOSTFLAGS_$(obj))
--- a/tools/perf/pmu-events/jevents.c
+++ b/tools/perf/pmu-events/jevents.c
@@ -35,6 +35,7 @@
#include <stdlib.h>
#include <errno.h>
#include <string.h>
+#include <strings.h>
#include <ctype.h>
#include <unistd.h>
#include <stdarg.h>
--- a/tools/perf/pmu-events/json.c
+++ b/tools/perf/pmu-events/json.c
@@ -38,7 +38,6 @@
#include <unistd.h>
#include "jsmn.h"
#include "json.h"
-#include <linux/kernel.h>
static char *mapfile(const char *fn, size_t *size)

65
target/linux/generic/hack-4.14/212-byteshift_portability.patch Normal file
View file

@ -0,0 +1,65 @@
From 48232d3d931c95953ce2ddfe7da7bb164aef6a73 Mon Sep 17 00:00:00 2001
From: Felix Fietkau <nbd@nbd.name>
Date: Fri, 7 Jul 2017 17:03:16 +0200
Subject: linux-3.6: fix portability of some includes files in tools/ used on the host
lede-commit: 6040b1d29ab1f047c5e49b748abcb6a3196add28
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
tools/include/tools/be_byteshift.h | 4 ++++
tools/include/tools/le_byteshift.h | 4 ++++
tools/include/tools/linux_types.h | 22 ++++++++++++++++++++++
3 files changed, 30 insertions(+)
create mode 100644 tools/include/tools/linux_types.h
--- a/tools/include/tools/be_byteshift.h
+++ b/tools/include/tools/be_byteshift.h
@@ -2,6 +2,10 @@
#ifndef _TOOLS_BE_BYTESHIFT_H
#define _TOOLS_BE_BYTESHIFT_H
+#ifndef __linux__
+#include "linux_types.h"
+#endif
+
#include <stdint.h>
static inline uint16_t __get_unaligned_be16(const uint8_t *p)
--- a/tools/include/tools/le_byteshift.h
+++ b/tools/include/tools/le_byteshift.h
@@ -2,6 +2,10 @@
#ifndef _TOOLS_LE_BYTESHIFT_H
#define _TOOLS_LE_BYTESHIFT_H
+#ifndef __linux__
+#include "linux_types.h"
+#endif
+
#include <stdint.h>
static inline uint16_t __get_unaligned_le16(const uint8_t *p)
--- /dev/null
+++ b/tools/include/tools/linux_types.h
@@ -0,0 +1,22 @@
+#ifndef __LINUX_TYPES_H
+#define __LINUX_TYPES_H
+
+#include <stdint.h>
+
+typedef uint8_t __u8;
+typedef uint8_t __be8;
+typedef uint8_t __le8;
+
+typedef uint16_t __u16;
+typedef uint16_t __be16;
+typedef uint16_t __le16;
+
+typedef uint32_t __u32;
+typedef uint32_t __be32;
+typedef uint32_t __le32;
+
+typedef uint64_t __u64;
+typedef uint64_t __be64;
+typedef uint64_t __le64;
+
+#endif

24
target/linux/generic/hack-4.14/214-spidev_h_portability.patch Normal file
View file

@ -0,0 +1,24 @@
From be9be95ff10e16a5b4ad36f903978d0cc5747024 Mon Sep 17 00:00:00 2001
From: Felix Fietkau <nbd@nbd.name>
Date: Fri, 7 Jul 2017 17:04:08 +0200
Subject: kernel: fix linux/spi/spidev.h portability issues with musl
Felix will try to get this define included into musl
lede-commit: 795e7cf60de19e7a076a46874fab7bb88b43bbff
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
include/uapi/linux/spi/spidev.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
--- a/include/uapi/linux/spi/spidev.h
+++ b/include/uapi/linux/spi/spidev.h
@@ -113,7 +113,7 @@ struct spi_ioc_transfer {
/* not all platforms use <asm-generic/ioctl.h> or _IOC_TYPECHECK() ... */
#define SPI_MSGSIZE(N) \
- ((((N)*(sizeof (struct spi_ioc_transfer))) < (1 << _IOC_SIZEBITS)) \
+ ((((N)*(sizeof (struct spi_ioc_transfer))) < (1 << 13)) \
? ((N)*(sizeof (struct spi_ioc_transfer))) : 0)
#define SPI_IOC_MESSAGE(N) _IOW(SPI_IOC_MAGIC, 0, char[SPI_MSGSIZE(N)])

258
target/linux/generic/hack-4.14/220-gc_sections.patch Normal file
View file

@ -0,0 +1,258 @@
From e3d8676f5722b7622685581e06e8f53e6138e3ab Mon Sep 17 00:00:00 2001
From: Felix Fietkau <nbd@nbd.name>
Date: Sat, 15 Jul 2017 23:42:36 +0200
Subject: use -ffunction-sections, -fdata-sections and --gc-sections
In combination with kernel symbol export stripping this significantly reduces
the kernel image size. Used on both ARM and MIPS architectures.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Jonas Gorski <jogo@openwrt.org>
Signed-off-by: Gabor Juhos <juhosg@openwrt.org>
---
Makefile | 10 +++----
arch/arm/Kconfig | 1 +
arch/arm/boot/compressed/Makefile | 1 +
arch/arm/kernel/vmlinux.lds.S | 26 ++++++++--------
arch/mips/Kconfig | 1 +
arch/mips/kernel/vmlinux.lds.S | 4 +--
include/asm-generic/vmlinux.lds.h | 63 ++++++++++++++++++++-------------------
7 files changed, 55 insertions(+), 51 deletions(-)
--- a/Makefile
+++ b/Makefile
@@ -410,6 +410,11 @@ KBUILD_AFLAGS_MODULE := -DMODULE
KBUILD_CFLAGS_MODULE := -DMODULE
KBUILD_LDFLAGS_MODULE = -T $(srctree)/scripts/module-common.lds $(if $(CONFIG_PROFILING),,-s)
+ifdef CONFIG_LD_DEAD_CODE_DATA_ELIMINATION
+KBUILD_CFLAGS_KERNEL += $(call cc-option,-ffunction-sections,)
+KBUILD_CFLAGS_KERNEL += $(call cc-option,-fdata-sections,)
+endif
+
# Read KERNELRELEASE from include/config/kernel.release (if it exists)
KERNELRELEASE = $(shell cat include/config/kernel.release 2> /dev/null)
KERNELVERSION = $(VERSION)$(if $(PATCHLEVEL),.$(PATCHLEVEL)$(if $(SUBLEVEL),.$(SUBLEVEL)))$(EXTRAVERSION)
@@ -783,11 +788,6 @@ ifdef CONFIG_DEBUG_SECTION_MISMATCH
KBUILD_CFLAGS += $(call cc-option, -fno-inline-functions-called-once)
endif
-ifdef CONFIG_LD_DEAD_CODE_DATA_ELIMINATION
-KBUILD_CFLAGS += $(call cc-option,-ffunction-sections,)
-KBUILD_CFLAGS += $(call cc-option,-fdata-sections,)
-endif
-
# arch Makefile may override CC so keep this after arch Makefile is included
NOSTDINC_FLAGS += -nostdinc -isystem $(shell $(CC) -print-file-name=include)
CHECKFLAGS += $(NOSTDINC_FLAGS)
--- a/arch/arm/Kconfig
+++ b/arch/arm/Kconfig
@@ -91,6 +91,7 @@ config ARM
select HAVE_UID16
select HAVE_VIRT_CPU_ACCOUNTING_GEN
select IRQ_FORCED_THREADING
+ select LD_DEAD_CODE_DATA_ELIMINATION
select MODULES_USE_ELF_REL
select NO_BOOTMEM
select OF_EARLY_FLATTREE if OF
--- a/arch/arm/boot/compressed/Makefile
+++ b/arch/arm/boot/compressed/Makefile
@@ -103,6 +103,7 @@ ifeq ($(CONFIG_FUNCTION_TRACER),y)
ORIG_CFLAGS := $(KBUILD_CFLAGS)
KBUILD_CFLAGS = $(subst -pg, , $(ORIG_CFLAGS))
endif
+KBUILD_CFLAGS_KERNEL := $(patsubst -f%-sections,,$(KBUILD_CFLAGS_KERNEL))
# -fstack-protector-strong triggers protection checks in this code,
# but it is being used too early to link to meaningful stack_chk logic.
--- a/arch/arm/kernel/vmlinux.lds.S
+++ b/arch/arm/kernel/vmlinux.lds.S
@@ -18,7 +18,7 @@
#define PROC_INFO \
. = ALIGN(4); \
VMLINUX_SYMBOL(__proc_info_begin) = .; \
- *(.proc.info.init) \
+ KEEP(*(.proc.info.init)) \
VMLINUX_SYMBOL(__proc_info_end) = .;
#define HYPERVISOR_TEXT \
@@ -29,11 +29,11 @@
#define IDMAP_TEXT \
ALIGN_FUNCTION(); \
VMLINUX_SYMBOL(__idmap_text_start) = .; \
- *(.idmap.text) \
+ KEEP(*(.idmap.text)) \
VMLINUX_SYMBOL(__idmap_text_end) = .; \
. = ALIGN(PAGE_SIZE); \
VMLINUX_SYMBOL(__hyp_idmap_text_start) = .; \
- *(.hyp.idmap.text) \
+ KEEP(*(.hyp.idmap.text)) \
VMLINUX_SYMBOL(__hyp_idmap_text_end) = .;
#ifdef CONFIG_HOTPLUG_CPU
@@ -106,7 +106,7 @@ SECTIONS
_stext = .; /* Text and read-only data */
IDMAP_TEXT
__exception_text_start = .;
- *(.exception.text)
+ KEEP(*(.exception.text))
__exception_text_end = .;
IRQENTRY_TEXT
SOFTIRQENTRY_TEXT
@@ -135,7 +135,7 @@ SECTIONS
__ex_table : AT(ADDR(__ex_table) - LOAD_OFFSET) {
__start___ex_table = .;
#ifdef CONFIG_MMU
- *(__ex_table)
+ KEEP(*(__ex_table))
#endif
__stop___ex_table = .;
}
@@ -147,12 +147,12 @@ SECTIONS
. = ALIGN(8);
.ARM.unwind_idx : {
__start_unwind_idx = .;
- *(.ARM.exidx*)
+ KEEP(*(.ARM.exidx*))
__stop_unwind_idx = .;
}
.ARM.unwind_tab : {
__start_unwind_tab = .;
- *(.ARM.extab*)
+ KEEP(*(.ARM.extab*))
__stop_unwind_tab = .;
}
#endif
@@ -172,14 +172,14 @@ SECTIONS
*/
__vectors_start = .;
.vectors 0xffff0000 : AT(__vectors_start) {
- *(.vectors)
+ KEEP(*(.vectors))
}
. = __vectors_start + SIZEOF(.vectors);
__vectors_end = .;
__stubs_start = .;
.stubs ADDR(.vectors) + 0x1000 : AT(__stubs_start) {
- *(.stubs)
+ KEEP(*(.stubs))
}
. = __stubs_start + SIZEOF(.stubs);
__stubs_end = .;
@@ -195,24 +195,24 @@ SECTIONS
}
.init.arch.info : {
__arch_info_begin = .;
- *(.arch.info.init)
+ KEEP(*(.arch.info.init))
__arch_info_end = .;
}
.init.tagtable : {
__tagtable_begin = .;
- *(.taglist.init)
+ KEEP(*(.taglist.init))
__tagtable_end = .;
}
#ifdef CONFIG_SMP_ON_UP
.init.smpalt : {
__smpalt_begin = .;
- *(.alt.smp.init)
+ KEEP(*(.alt.smp.init))
__smpalt_end = .;
}
#endif
.init.pv_table : {
__pv_table_begin = .;
- *(.pv_table)
+ KEEP(*(.pv_table))
__pv_table_end = .;
}
.init.data : {
--- a/arch/mips/Kconfig
+++ b/arch/mips/Kconfig
@@ -39,6 +39,7 @@ config MIPS
select HAVE_CBPF_JIT if (!64BIT && !CPU_MICROMIPS)
select HAVE_EBPF_JIT if (64BIT && !CPU_MICROMIPS)
select HAVE_CC_STACKPROTECTOR
+ select LD_DEAD_CODE_DATA_ELIMINATION
select HAVE_CONTEXT_TRACKING
select HAVE_COPY_THREAD_TLS
select HAVE_C_RECORDMCOUNT
--- a/arch/mips/kernel/vmlinux.lds.S
+++ b/arch/mips/kernel/vmlinux.lds.S
@@ -72,7 +72,7 @@ SECTIONS
/* Exception table for data bus errors */
__dbe_table : {
__start___dbe_table = .;
- *(__dbe_table)
+ KEEP(*(__dbe_table))
__stop___dbe_table = .;
}
@@ -123,7 +123,7 @@ SECTIONS
. = ALIGN(4);
.mips.machines.init : AT(ADDR(.mips.machines.init) - LOAD_OFFSET) {
__mips_machines_start = .;
- *(.mips.machines.init)
+ KEEP(*(.mips.machines.init))
__mips_machines_end = .;
}
--- a/include/asm-generic/vmlinux.lds.h
+++ b/include/asm-generic/vmlinux.lds.h
@@ -105,7 +105,7 @@
#ifdef CONFIG_FTRACE_MCOUNT_RECORD
#define MCOUNT_REC() . = ALIGN(8); \
VMLINUX_SYMBOL(__start_mcount_loc) = .; \
- *(__mcount_loc) \
+ KEEP(*(__mcount_loc)) \
VMLINUX_SYMBOL(__stop_mcount_loc) = .;
#else
#define MCOUNT_REC()
@@ -113,7 +113,7 @@
#ifdef CONFIG_TRACE_BRANCH_PROFILING
#define LIKELY_PROFILE() VMLINUX_SYMBOL(__start_annotated_branch_profile) = .; \
- *(_ftrace_annotated_branch) \
+ KEEP(*(_ftrace_annotated_branch)) \
VMLINUX_SYMBOL(__stop_annotated_branch_profile) = .;
#else
#define LIKELY_PROFILE()
@@ -121,7 +121,7 @@
#ifdef CONFIG_PROFILE_ALL_BRANCHES
#define BRANCH_PROFILE() VMLINUX_SYMBOL(__start_branch_profile) = .; \
- *(_ftrace_branch) \
+ KEEP(*(_ftrace_branch)) \
VMLINUX_SYMBOL(__stop_branch_profile) = .;
#else
#define BRANCH_PROFILE()
@@ -237,7 +237,8 @@
LIKELY_PROFILE() \
BRANCH_PROFILE() \
TRACE_PRINTKS() \
- TRACEPOINT_STR()
+ TRACEPOINT_STR() \
+ *(.data.[a-zA-Z_]*)
/*
* Data section helpers
@@ -496,7 +497,7 @@
#define ENTRY_TEXT \
ALIGN_FUNCTION(); \
VMLINUX_SYMBOL(__entry_text_start) = .; \
- *(.entry.text) \
+ KEEP(*(.entry.text)) \
VMLINUX_SYMBOL(__entry_text_end) = .;
#define IRQENTRY_TEXT \
@@ -603,7 +604,7 @@
. = ALIGN(sbss_align); \
.sbss : AT(ADDR(.sbss) - LOAD_OFFSET) { \
*(.dynsbss) \
- *(.sbss) \
+ *(.sbss .sbss.*) \
*(.scommon) \
}

101
target/linux/generic/hack-4.14/221-module_exports.patch Normal file
View file

@ -0,0 +1,101 @@
From b14784e7883390c20ed3ff904892255404a5914b Mon Sep 17 00:00:00 2001
From: Felix Fietkau <nbd@nbd.name>
Date: Fri, 7 Jul 2017 17:05:53 +0200
Subject: add an optional config option for stripping all unnecessary symbol exports from the kernel image
lede-commit: bb5a40c64b7c4f4848509fa0a6625055fc9e66cc
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
include/asm-generic/vmlinux.lds.h | 18 +++++++++++++++---
include/linux/export.h | 9 ++++++++-
scripts/Makefile.build | 2 +-
3 files changed, 24 insertions(+), 5 deletions(-)
--- a/include/asm-generic/vmlinux.lds.h
+++ b/include/asm-generic/vmlinux.lds.h
@@ -54,6 +54,16 @@
#define LOAD_OFFSET 0
#endif
+#ifndef SYMTAB_KEEP
+#define SYMTAB_KEEP KEEP(*(SORT(___ksymtab+*)))
+#define SYMTAB_KEEP_GPL KEEP(*(SORT(___ksymtab_gpl+*)))
+#endif
+
+#ifndef SYMTAB_DISCARD
+#define SYMTAB_DISCARD
+#define SYMTAB_DISCARD_GPL
+#endif
+
#include <linux/export.h>
/* Align . to a 8 byte boundary equals to maximum function alignment. */
@@ -341,14 +351,14 @@
/* Kernel symbol table: Normal symbols */ \
__ksymtab : AT(ADDR(__ksymtab) - LOAD_OFFSET) { \
VMLINUX_SYMBOL(__start___ksymtab) = .; \
- KEEP(*(SORT(___ksymtab+*))) \
+ SYMTAB_KEEP \
VMLINUX_SYMBOL(__stop___ksymtab) = .; \
} \
\
/* Kernel symbol table: GPL-only symbols */ \
__ksymtab_gpl : AT(ADDR(__ksymtab_gpl) - LOAD_OFFSET) { \
VMLINUX_SYMBOL(__start___ksymtab_gpl) = .; \
- KEEP(*(SORT(___ksymtab_gpl+*))) \
+ SYMTAB_KEEP_GPL \
VMLINUX_SYMBOL(__stop___ksymtab_gpl) = .; \
} \
\
@@ -410,7 +420,7 @@
\
/* Kernel symbol table: strings */ \
__ksymtab_strings : AT(ADDR(__ksymtab_strings) - LOAD_OFFSET) { \
- *(__ksymtab_strings) \
+ *(__ksymtab_strings+*) \
} \
\
/* __*init sections */ \
@@ -793,6 +803,8 @@
EXIT_TEXT \
EXIT_DATA \
EXIT_CALL \
+ SYMTAB_DISCARD \
+ SYMTAB_DISCARD_GPL \
*(.discard) \
*(.discard.*) \
}
--- a/include/linux/export.h
+++ b/include/linux/export.h
@@ -60,12 +60,19 @@ extern struct module __this_module;
#define __CRC_SYMBOL(sym, sec)
#endif
+#ifdef MODULE
+#define __EXPORT_SUFFIX(sym)
+#else
+#define __EXPORT_SUFFIX(sym) "+" #sym
+#endif
+
/* For every exported symbol, place a struct in the __ksymtab section */
#define ___EXPORT_SYMBOL(sym, sec) \
extern typeof(sym) sym; \
__CRC_SYMBOL(sym, sec) \
static const char __kstrtab_##sym[] \
- __attribute__((section("__ksymtab_strings"), aligned(1))) \
+ __attribute__((section("__ksymtab_strings" \
+ __EXPORT_SUFFIX(sym)), aligned(1))) \
= VMLINUX_SYMBOL_STR(sym); \
static const struct kernel_symbol __ksymtab_##sym \
__used \
--- a/scripts/Makefile.build
+++ b/scripts/Makefile.build
@@ -420,7 +420,7 @@ targets += $(extra-y) $(MAKECMDGOALS) $(
# Linker scripts preprocessor (.lds.S -> .lds)
# ---------------------------------------------------------------------------
quiet_cmd_cpp_lds_S = LDS $@
- cmd_cpp_lds_S = $(CPP) $(cpp_flags) -P -U$(ARCH) \
+ cmd_cpp_lds_S = $(CPP) $(EXTRA_LDSFLAGS) $(cpp_flags) -P -U$(ARCH) \
-D__ASSEMBLY__ -DLINKER_SCRIPT -o $@ $<
$(obj)/%.lds: $(src)/%.lds.S FORCE

71
target/linux/generic/hack-4.14/230-openwrt_lzma_options.patch Normal file
View file

@ -0,0 +1,71 @@
From b3d00b452467f621317953d9e4c6f9ae8dcfd271 Mon Sep 17 00:00:00 2001
From: Imre Kaloz <kaloz@openwrt.org>
Date: Fri, 7 Jul 2017 17:06:55 +0200
Subject: use the openwrt lzma options for now
lede-commit: 548de949f392049420a6a1feeef118b30ab8ea8c
Signed-off-by: Imre Kaloz <kaloz@openwrt.org>
---
lib/decompress.c | 1 +
scripts/Makefile.lib | 2 +-
scripts/gen_initramfs_list.sh | 10 +++++-----
3 files changed, 7 insertions(+), 6 deletions(-)
--- a/lib/decompress.c
+++ b/lib/decompress.c
@@ -49,6 +49,7 @@ static const struct compress_format comp
{ {0x1f, 0x9e}, "gzip", gunzip },
{ {0x42, 0x5a}, "bzip2", bunzip2 },
{ {0x5d, 0x00}, "lzma", unlzma },
+ { {0x6d, 0x00}, "lzma-openwrt", unlzma },
{ {0xfd, 0x37}, "xz", unxz },
{ {0x89, 0x4c}, "lzo", unlzo },
{ {0x02, 0x21}, "lz4", unlz4 },
--- a/scripts/Makefile.lib
+++ b/scripts/Makefile.lib
@@ -348,7 +348,7 @@ cmd_bzip2 = (cat $(filter-out FORCE,$^)
quiet_cmd_lzma = LZMA $@
cmd_lzma = (cat $(filter-out FORCE,$^) | \
- lzma -9 && $(call size_append, $(filter-out FORCE,$^))) > $@ || \
+ lzma e -d20 -lc1 -lp2 -pb2 -eos -si -so && $(call size_append, $(filter-out FORCE,$^))) > $@ || \
(rm -f $@ ; false)
quiet_cmd_lzo = LZO $@
--- a/scripts/gen_initramfs_list.sh
+++ b/scripts/gen_initramfs_list.sh
@@ -229,7 +229,7 @@ cpio_list=
output="/dev/stdout"
output_file=""
is_cpio_compressed=
-compr="gzip -n -9 -f"
+compr="gzip -n -9 -f -"
arg="$1"
case "$arg" in
@@ -245,13 +245,13 @@ case "$arg" in
output=${cpio_list}
echo "$output_file" | grep -q "\.gz$" \
&& [ -x "`which gzip 2> /dev/null`" ] \
- && compr="gzip -n -9 -f"
+ && compr="gzip -n -9 -f -"
echo "$output_file" | grep -q "\.bz2$" \
&& [ -x "`which bzip2 2> /dev/null`" ] \
- && compr="bzip2 -9 -f"
+ && compr="bzip2 -9 -f -"
echo "$output_file" | grep -q "\.lzma$" \
&& [ -x "`which lzma 2> /dev/null`" ] \
- && compr="lzma -9 -f"
+ && compr="lzma e -d20 -lc1 -lp2 -pb2 -eos -si -so"
echo "$output_file" | grep -q "\.xz$" \
&& [ -x "`which xz 2> /dev/null`" ] \
&& compr="xz --check=crc32 --lzma2=dict=1MiB"
@@ -320,7 +320,7 @@ if [ ! -z ${output_file} ]; then
if [ "${is_cpio_compressed}" = "compressed" ]; then
cat ${cpio_tfile} > ${output_file}
else
- (cat ${cpio_tfile} | ${compr} - > ${output_file}) \
+ (cat ${cpio_tfile} | ${compr} > ${output_file}) \
|| (rm -f ${output_file} ; false)
fi
[ -z ${cpio_file} ] && rm ${cpio_tfile}

27
target/linux/generic/hack-4.14/250-netfilter_depends.patch Normal file
View file

@ -0,0 +1,27 @@
From: Felix Fietkau <nbd@nbd.name>
Subject: hack: net: remove bogus netfilter dependencies
lede-commit: 589d2a377dee27d206fc3725325309cf649e4df6
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
net/netfilter/Kconfig | 2 --
1 file changed, 2 deletions(-)
--- a/net/netfilter/Kconfig
+++ b/net/netfilter/Kconfig
@@ -223,7 +223,6 @@ config NF_CONNTRACK_FTP
config NF_CONNTRACK_H323
tristate "H.323 protocol support"
- depends on IPV6 || IPV6=n
depends on NETFILTER_ADVANCED
help
H.323 is a VoIP signalling protocol from ITU-T. As one of the most
@@ -1012,7 +1011,6 @@ config NETFILTER_XT_TARGET_SECMARK
config NETFILTER_XT_TARGET_TCPMSS
tristate '"TCPMSS" target support'
- depends on IPV6 || IPV6=n
default m if NETFILTER_ADVANCED=n
---help---
This option adds a `TCPMSS' target, which allows you to alter the

197
target/linux/generic/hack-4.14/251-sound_kconfig.patch Normal file
View file

@ -0,0 +1,197 @@
From da3c50704f14132f4adf80d48e9a4cd5d46e54c9 Mon Sep 17 00:00:00 2001
From: John Crispin <john@phrozen.org>
Date: Fri, 7 Jul 2017 17:09:21 +0200
Subject: kconfig: owrt specifc dependencies
Signed-off-by: John Crispin <john@phrozen.org>
---
crypto/Kconfig | 10 +++++-----
drivers/bcma/Kconfig | 1 +
drivers/ssb/Kconfig | 3 ++-
lib/Kconfig | 8 ++++----
net/netfilter/Kconfig | 2 +-
net/wireless/Kconfig | 17 ++++++++++-------
sound/core/Kconfig | 4 ++--
7 files changed, 25 insertions(+), 20 deletions(-)
--- a/crypto/Kconfig
+++ b/crypto/Kconfig
@@ -33,7 +33,7 @@ config CRYPTO_FIPS
this is.
config CRYPTO_ALGAPI
- tristate
+ tristate "ALGAPI"
select CRYPTO_ALGAPI2
help
This option provides the API for cryptographic algorithms.
@@ -42,7 +42,7 @@ config CRYPTO_ALGAPI2
tristate
config CRYPTO_AEAD
- tristate
+ tristate "AEAD"
select CRYPTO_AEAD2
select CRYPTO_ALGAPI
@@ -53,7 +53,7 @@ config CRYPTO_AEAD2
select CRYPTO_RNG2
config CRYPTO_BLKCIPHER
- tristate
+ tristate "BLKCIPHER"
select CRYPTO_BLKCIPHER2
select CRYPTO_ALGAPI
@@ -64,7 +64,7 @@ config CRYPTO_BLKCIPHER2
select CRYPTO_WORKQUEUE
config CRYPTO_HASH
- tristate
+ tristate "HASH"
select CRYPTO_HASH2
select CRYPTO_ALGAPI
@@ -73,7 +73,7 @@ config CRYPTO_HASH2
select CRYPTO_ALGAPI2
config CRYPTO_RNG
- tristate
+ tristate "RNG"
select CRYPTO_RNG2
select CRYPTO_ALGAPI
--- a/drivers/bcma/Kconfig
+++ b/drivers/bcma/Kconfig
@@ -15,6 +15,7 @@ menuconfig BCMA
config BCMA_BLOCKIO
bool
depends on BCMA
+ default y
config BCMA_HOST_PCI_POSSIBLE
bool
--- a/drivers/ssb/Kconfig
+++ b/drivers/ssb/Kconfig
@@ -29,6 +29,7 @@ config SSB_SPROM
config SSB_BLOCKIO
bool
depends on SSB
+ default y
config SSB_PCIHOST_POSSIBLE
bool
@@ -49,7 +50,7 @@ config SSB_PCIHOST
config SSB_B43_PCI_BRIDGE
bool
depends on SSB_PCIHOST
- default n
+ default y
config SSB_PCMCIAHOST_POSSIBLE
bool
--- a/lib/Kconfig
+++ b/lib/Kconfig
@@ -358,16 +358,16 @@ config BCH_CONST_T
# Textsearch support is select'ed if needed
#
config TEXTSEARCH
- bool
+ boolean "Textsearch support"
config TEXTSEARCH_KMP
- tristate
+ tristate "Textsearch KMP"
config TEXTSEARCH_BM
- tristate
+ tristate "Textsearch BM"
config TEXTSEARCH_FSM
- tristate
+ tristate "Textsearch FSM"
config BTREE
bool
--- a/net/netfilter/Kconfig
+++ b/net/netfilter/Kconfig
@@ -10,7 +10,7 @@ config NETFILTER_INGRESS
infrastructure.
config NETFILTER_NETLINK
- tristate
+ tristate "Netfilter NFNETLINK interface"
config NETFILTER_NETLINK_ACCT
tristate "Netfilter NFACCT over NFNETLINK interface"
--- a/net/wireless/Kconfig
+++ b/net/wireless/Kconfig
@@ -1,5 +1,5 @@
config WIRELESS_EXT
- bool
+ bool "Wireless extensions"
config WEXT_CORE
def_bool y
@@ -11,10 +11,10 @@ config WEXT_PROC
depends on WEXT_CORE
config WEXT_SPY
- bool
+ bool "WEXT_SPY"
config WEXT_PRIV
- bool
+ bool "WEXT_PRIV"
config CFG80211
tristate "cfg80211 - wireless configuration API"
@@ -188,7 +188,7 @@ config CFG80211_WEXT_EXPORT
wext compatibility symbols to be exported.
config LIB80211
- tristate
+ tristate "LIB80211"
default n
help
This options enables a library of common routines used
@@ -197,13 +197,16 @@ config LIB80211
Drivers should select this themselves if needed.
config LIB80211_CRYPT_WEP
- tristate
+ tristate "LIB80211_CRYPT_WEP"
+ select LIB80211
config LIB80211_CRYPT_CCMP
- tristate
+ tristate "LIB80211_CRYPT_CCMP"
+ select LIB80211
config LIB80211_CRYPT_TKIP
- tristate
+ tristate "LIB80211_CRYPT_TKIP"
+ select LIB80211
config LIB80211_DEBUG
bool "lib80211 debugging messages"
--- a/sound/core/Kconfig
+++ b/sound/core/Kconfig
@@ -16,7 +16,7 @@ config SND_DMAENGINE_PCM
tristate
config SND_HWDEP
- tristate
+ tristate "Sound hardware support"
config SND_SEQ_DEVICE
tristate
@@ -26,7 +26,7 @@ config SND_RAWMIDI
select SND_SEQ_DEVICE if SND_SEQUENCER != n
config SND_COMPRESS_OFFLOAD
- tristate
+ tristate "Compression offloading support"
config SND_JACK
bool

109
target/linux/generic/hack-4.14/259-regmap_dynamic.patch Normal file
View file

@ -0,0 +1,109 @@
From 811d9e2268a62b830cfe93cd8bc929afcb8b198b Mon Sep 17 00:00:00 2001
From: Felix Fietkau <nbd@nbd.name>
Date: Sat, 15 Jul 2017 21:12:38 +0200
Subject: kernel: move regmap bloat out of the kernel image if it is only being used in modules
lede-commit: 96f39119815028073583e4fca3a9c5fe9141e998
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
drivers/base/regmap/Kconfig | 15 ++++++++++-----
drivers/base/regmap/Makefile | 12 ++++++++----
drivers/base/regmap/regmap.c | 3 +++
include/linux/regmap.h | 2 +-
4 files changed, 22 insertions(+), 10 deletions(-)
--- a/drivers/base/regmap/Kconfig
+++ b/drivers/base/regmap/Kconfig
@@ -4,9 +4,8 @@
# subsystems should select the appropriate symbols.
config REGMAP
- default y if (REGMAP_I2C || REGMAP_SPI || REGMAP_SPMI || REGMAP_W1 || REGMAP_AC97 || REGMAP_MMIO || REGMAP_IRQ)
select IRQ_DOMAIN if REGMAP_IRQ
- bool
+ tristate "Regmap"
config REGCACHE_COMPRESSED
select LZO_COMPRESS
@@ -17,23 +16,30 @@ config REGMAP_AC97
tristate
config REGMAP_I2C
- tristate
+ tristate "Regmap I2C"
+ select REGMAP
depends on I2C
config REGMAP_SPI
- tristate
+ tristate "Regmap SPI"
+ select REGMAP
+ depends on SPI_MASTER
depends on SPI
config REGMAP_SPMI
+ select REGMAP
tristate
depends on SPMI
config REGMAP_W1
+ select REGMAP
tristate
depends on W1
config REGMAP_MMIO
- tristate
+ tristate "Regmap MMIO"
+ select REGMAP
config REGMAP_IRQ
+ select REGMAP
bool
--- a/drivers/base/regmap/Makefile
+++ b/drivers/base/regmap/Makefile
@@ -2,10 +2,14 @@
# For include/trace/define_trace.h to include trace.h
CFLAGS_regmap.o := -I$(src)
-obj-$(CONFIG_REGMAP) += regmap.o regcache.o
-obj-$(CONFIG_REGMAP) += regcache-rbtree.o regcache-flat.o
-obj-$(CONFIG_REGCACHE_COMPRESSED) += regcache-lzo.o
-obj-$(CONFIG_DEBUG_FS) += regmap-debugfs.o
+regmap-core-objs = regmap.o regcache.o regcache-rbtree.o regcache-flat.o
+ifdef CONFIG_DEBUG_FS
+regmap-core-objs += regmap-debugfs.o
+endif
+ifdef CONFIG_REGCACHE_COMPRESSED
+regmap-core-objs += regcache-lzo.o
+endif
+obj-$(CONFIG_REGMAP) += regmap-core.o
obj-$(CONFIG_REGMAP_AC97) += regmap-ac97.o
obj-$(CONFIG_REGMAP_I2C) += regmap-i2c.o
obj-$(CONFIG_REGMAP_SPI) += regmap-spi.o
--- a/drivers/base/regmap/regmap.c
+++ b/drivers/base/regmap/regmap.c
@@ -13,6 +13,7 @@
#include <linux/device.h>
#include <linux/slab.h>
#include <linux/export.h>
+#include <linux/module.h>
#include <linux/mutex.h>
#include <linux/err.h>
#include <linux/of.h>
@@ -2926,3 +2927,5 @@ static int __init regmap_initcall(void)
return 0;
}
postcore_initcall(regmap_initcall);
+
+MODULE_LICENSE("GPL");
--- a/include/linux/regmap.h
+++ b/include/linux/regmap.h
@@ -139,7 +139,7 @@ struct reg_sequence {
pollret ?: ((cond) ? 0 : -ETIMEDOUT); \
})
-#ifdef CONFIG_REGMAP
+#if IS_ENABLED(CONFIG_REGMAP)
enum regmap_endian {
/* Unspecified -> 0 -> Backwards compatible default */

60
target/linux/generic/hack-4.14/260-crypto_test_dependencies.patch Normal file
View file

@ -0,0 +1,60 @@
From fd1799b0bf5efa46dd3e6dfbbf3955564807e508 Mon Sep 17 00:00:00 2001
From: Felix Fietkau <nbd@nbd.name>
Date: Fri, 7 Jul 2017 17:12:51 +0200
Subject: kernel: prevent cryptomgr from pulling in useless extra dependencies for tests that are not run
Reduces kernel size after LZMA by about 5k on MIPS
lede-commit: 044c316167e076479a344c59905e5b435b84a77f
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
crypto/Kconfig | 13 ++++++-------
crypto/algboss.c | 4 ++++
2 files changed, 10 insertions(+), 7 deletions(-)
--- a/crypto/Kconfig
+++ b/crypto/Kconfig
@@ -143,13 +143,13 @@ config CRYPTO_MANAGER
cbc(aes).
config CRYPTO_MANAGER2
- def_tristate CRYPTO_MANAGER || (CRYPTO_MANAGER!=n && CRYPTO_ALGAPI=y)
- select CRYPTO_AEAD2
- select CRYPTO_HASH2
- select CRYPTO_BLKCIPHER2
- select CRYPTO_AKCIPHER2
- select CRYPTO_KPP2
- select CRYPTO_ACOMP2
+ def_tristate CRYPTO_MANAGER || (CRYPTO_MANAGER!=n && CRYPTO_ALGAPI=y && !CRYPTO_MANAGER_DISABLE_TESTS)
+ select CRYPTO_AEAD2 if !CRYPTO_MANAGER_DISABLE_TESTS
+ select CRYPTO_HASH2 if !CRYPTO_MANAGER_DISABLE_TESTS
+ select CRYPTO_BLKCIPHER2 if !CRYPTO_MANAGER_DISABLE_TESTS
+ select CRYPTO_AKCIPHER2 if !CRYPTO_MANAGER_DISABLE_TESTS
+ select CRYPTO_KPP2 if !CRYPTO_MANAGER_DISABLE_TESTS
+ select CRYPTO_ACOMP2 if !CRYPTO_MANAGER_DISABLE_TESTS
config CRYPTO_USER
tristate "Userspace cryptographic algorithm configuration"
@@ -162,7 +162,6 @@ config CRYPTO_USER
config CRYPTO_MANAGER_DISABLE_TESTS
bool "Disable run-time self tests"
default y
- depends on CRYPTO_MANAGER2
help
Disable run-time self tests that normally take place at
algorithm registration.
--- a/crypto/algboss.c
+++ b/crypto/algboss.c
@@ -248,8 +248,12 @@ static int cryptomgr_schedule_test(struc
type = alg->cra_flags;
/* Do not test internal algorithms. */
+#ifdef CONFIG_CRYPTO_MANAGER_DISABLE_TESTS
+ type |= CRYPTO_ALG_TESTED;
+#else
if (type & CRYPTO_ALG_INTERNAL)
type |= CRYPTO_ALG_TESTED;
+#endif
param->type = type;

84
target/linux/generic/hack-4.14/280-rfkill-stubs.patch Normal file
View file

@ -0,0 +1,84 @@
From 236c1acdfef5958010ac9814a9872e0a46fd78ee Mon Sep 17 00:00:00 2001
From: John Crispin <john@phrozen.org>
Date: Fri, 7 Jul 2017 17:13:44 +0200
Subject: rfkill: add fake rfkill support
allow building of modules depending on RFKILL even if RFKILL is not enabled.
Signed-off-by: John Crispin <john@phrozen.org>
---
include/linux/rfkill.h | 2 +-
net/Makefile | 2 +-
net/rfkill/Kconfig | 14 +++++++++-----
net/rfkill/Makefile | 2 +-
4 files changed, 12 insertions(+), 8 deletions(-)
--- a/include/linux/rfkill.h
+++ b/include/linux/rfkill.h
@@ -64,7 +64,7 @@ struct rfkill_ops {
int (*set_block)(void *data, bool blocked);
};
-#if defined(CONFIG_RFKILL) || defined(CONFIG_RFKILL_MODULE)
+#if defined(CONFIG_RFKILL_FULL) || defined(CONFIG_RFKILL_FULL_MODULE)
/**
* rfkill_alloc - allocate rfkill structure
* @name: name of the struct -- the string is not copied internally
--- a/net/Makefile
+++ b/net/Makefile
@@ -53,7 +53,7 @@ obj-$(CONFIG_TIPC) += tipc/
obj-$(CONFIG_NETLABEL) += netlabel/
obj-$(CONFIG_IUCV) += iucv/
obj-$(CONFIG_SMC) += smc/
-obj-$(CONFIG_RFKILL) += rfkill/
+obj-$(CONFIG_RFKILL_FULL) += rfkill/
obj-$(CONFIG_NET_9P) += 9p/
obj-$(CONFIG_CAIF) += caif/
ifneq ($(CONFIG_DCB),)
--- a/net/rfkill/Kconfig
+++ b/net/rfkill/Kconfig
@@ -1,7 +1,11 @@
#
# RF switch subsystem configuration
#
-menuconfig RFKILL
+config RFKILL
+ bool
+ default y
+
+menuconfig RFKILL_FULL
tristate "RF switch subsystem support"
help
Say Y here if you want to have control over RF switches
@@ -13,19 +17,19 @@ menuconfig RFKILL
# LED trigger support
config RFKILL_LEDS
bool
- depends on RFKILL
+ depends on RFKILL_FULL
depends on LEDS_TRIGGERS = y || RFKILL = LEDS_TRIGGERS
default y
config RFKILL_INPUT
bool "RF switch input support" if EXPERT
- depends on RFKILL
+ depends on RFKILL_FULL
depends on INPUT = y || RFKILL = INPUT
default y if !EXPERT
config RFKILL_GPIO
tristate "GPIO RFKILL driver"
- depends on RFKILL
+ depends on RFKILL_FULL
depends on GPIOLIB || COMPILE_TEST
default n
help
--- a/net/rfkill/Makefile
+++ b/net/rfkill/Makefile
@@ -4,5 +4,5 @@
rfkill-y += core.o
rfkill-$(CONFIG_RFKILL_INPUT) += input.o
-obj-$(CONFIG_RFKILL) += rfkill.o
+obj-$(CONFIG_RFKILL_FULL) += rfkill.o
obj-$(CONFIG_RFKILL_GPIO) += rfkill-gpio.o

38
target/linux/generic/hack-4.14/301-mips_image_cmdline_hack.patch Normal file
View file

@ -0,0 +1,38 @@
From: John Crispin <john@phrozen.org>
Subject: hack: kernel: add generic image_cmdline hack to MIPS targets
lede-commit: d59f5b3a987a48508257a0ddbaeadc7909f9f976
Signed-off-by: Gabor Juhos <juhosg@openwrt.org>
---
arch/mips/Kconfig | 4 ++++
arch/mips/kernel/head.S | 6 ++++++
2 files changed, 10 insertions(+)
--- a/arch/mips/Kconfig
+++ b/arch/mips/Kconfig
@@ -1157,6 +1157,10 @@ config SYNC_R4K
config MIPS_MACHINE
def_bool n
+config IMAGE_CMDLINE_HACK
+ bool "OpenWrt specific image command line hack"
+ default n
+
config NO_IOPORT_MAP
def_bool n
--- a/arch/mips/kernel/head.S
+++ b/arch/mips/kernel/head.S
@@ -79,6 +79,12 @@ FEXPORT(__kernel_entry)
j kernel_entry
#endif
+#ifdef CONFIG_IMAGE_CMDLINE_HACK
+ .ascii "CMDLINE:"
+EXPORT(__image_cmdline)
+ .fill 0x400
+#endif /* CONFIG_IMAGE_CMDLINE_HACK */
+
__REF
NESTED(kernel_entry, 16, sp) # kernel entry point

38
target/linux/generic/hack-4.14/321-powerpc_crtsavres_prereq.patch Normal file
View file

@ -0,0 +1,38 @@
From 107c0964cb8db7ca28ac5199426414fdab3c274d Mon Sep 17 00:00:00 2001
From: "Alexandros C. Couloumbis" <alex@ozo.com>
Date: Fri, 7 Jul 2017 17:14:51 +0200
Subject: hack: arch: powerpc: drop register save/restore library from modules
Upstream GCC uses a libgcc function for saving/restoring registers. This
makes the code bigger, and upstream kernels need to carry that function
for every single kernel module. Our GCC is patched to avoid those
references, so we can drop the extra bloat for modules.
lede-commit: e8e1084654f50904e6bf77b70b2de3f137d7b3ec
Signed-off-by: Alexandros C. Couloumbis <alex@ozo.com>
---
arch/powerpc/Makefile | 1 -
1 file changed, 1 deletion(-)
--- a/arch/powerpc/Makefile
+++ b/arch/powerpc/Makefile
@@ -59,19 +59,6 @@ machine-$(CONFIG_PPC64) += 64
machine-$(CONFIG_CPU_LITTLE_ENDIAN) += le
UTS_MACHINE := $(subst $(space),,$(machine-y))
-# XXX This needs to be before we override LD below
-ifdef CONFIG_PPC32
-KBUILD_LDFLAGS_MODULE += arch/powerpc/lib/crtsavres.o
-else
-ifeq ($(call ld-ifversion, -ge, 225000000, y),y)
-# Have the linker provide sfpr if possible.
-# There is a corresponding test in arch/powerpc/lib/Makefile
-KBUILD_LDFLAGS_MODULE += --save-restore-funcs
-else
-KBUILD_LDFLAGS_MODULE += arch/powerpc/lib/crtsavres.o
-endif
-endif
-
ifeq ($(CONFIG_CPU_LITTLE_ENDIAN),y)
override LD += -EL
LDEMULATION := lppc

1040
target/linux/generic/hack-4.14/531-debloat_lzma.patch Normal file
View file

File diff suppressed because it is too large Load diff

32
target/linux/generic/hack-4.14/640-bridge-only-accept-EAP-locally.patch Normal file
View file

@ -0,0 +1,32 @@
From c6905cfdeb31a5c049db3da434b10fa0d3e83569 Mon Sep 17 00:00:00 2001
From: Felix Fietkau <nbd@nbd.name>
Date: Fri, 7 Jul 2017 17:18:54 +0200
Subject: bridge: only accept EAP locally
When bridging, do not forward EAP frames to other ports, only deliver
them locally, regardless of the state.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
net/bridge/br_input.c | 7 +++++--
1 file changed, 5 insertions(+), 2 deletions(-)
--- a/net/bridge/br_input.c
+++ b/net/bridge/br_input.c
@@ -166,11 +166,14 @@ int br_handle_frame_finish(struct net *n
}
}
+ BR_INPUT_SKB_CB(skb)->brdev = br->dev;
+
+ if (skb->protocol == htons(ETH_P_PAE))
+ return br_pass_frame_up(skb);
+
if (p->state == BR_STATE_LEARNING)
goto drop;
- BR_INPUT_SKB_CB(skb)->brdev = br->dev;
-
if (IS_ENABLED(CONFIG_INET) && skb->protocol == htons(ETH_P_ARP))
br_do_proxy_arp(skb, br, vid, p);

76
target/linux/generic/hack-4.14/641-bridge_port_isolate.patch Normal file
View file

@ -0,0 +1,76 @@
From e988390850731aa1697ed09d47b0932fac1af175 Mon Sep 17 00:00:00 2001
From: Felix Fietkau <nbd@nbd.name>
Date: Fri, 7 Jul 2017 17:20:03 +0200
Subject: bridge: port isolate
Isolating individual bridge ports
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
include/linux/if_bridge.h | 1 +
net/bridge/br_forward.c | 5 +++++
net/bridge/br_input.c | 3 +++
net/bridge/br_sysfs_if.c | 2 ++
4 files changed, 11 insertions(+)
--- a/include/linux/if_bridge.h
+++ b/include/linux/if_bridge.h
@@ -49,6 +49,7 @@ struct br_ip_list {
#define BR_MULTICAST_TO_UNICAST BIT(12)
#define BR_VLAN_TUNNEL BIT(13)
#define BR_BCAST_FLOOD BIT(14)
+#define BR_ISOLATE_MODE BIT(15)
#define BR_DEFAULT_AGEING_TIME (300 * HZ)
--- a/net/bridge/br_forward.c
+++ b/net/bridge/br_forward.c
@@ -141,6 +141,9 @@ static int deliver_clone(const struct ne
void br_forward(const struct net_bridge_port *to,
struct sk_buff *skb, bool local_rcv, bool local_orig)
{
+ if (to->flags & BR_ISOLATE_MODE)
+ to = NULL;
+
if (to && should_deliver(to, skb)) {
if (local_rcv)
deliver_clone(to, skb, local_orig);
@@ -183,6 +186,8 @@ void br_flood(struct net_bridge *br, str
struct net_bridge_port *p;
list_for_each_entry_rcu(p, &br->port_list, list) {
+ if (!local_orig && (p->flags & BR_ISOLATE_MODE))
+ continue;
/* Do not flood unicast traffic to ports that turn it off, nor
* other traffic if flood off, except for traffic we originate
*/
--- a/net/bridge/br_input.c
+++ b/net/bridge/br_input.c
@@ -177,6 +177,9 @@ int br_handle_frame_finish(struct net *n
if (IS_ENABLED(CONFIG_INET) && skb->protocol == htons(ETH_P_ARP))
br_do_proxy_arp(skb, br, vid, p);
+ if (p->flags & BR_ISOLATE_MODE)
+ return br_pass_frame_up(skb);
+
switch (pkt_type) {
case BR_PKT_MULTICAST:
mdst = br_mdb_get(br, skb, vid);
--- a/net/bridge/br_sysfs_if.c
+++ b/net/bridge/br_sysfs_if.c
@@ -174,6 +174,7 @@ BRPORT_ATTR_FLAG(proxyarp, BR_PROXYARP);
BRPORT_ATTR_FLAG(proxyarp_wifi, BR_PROXYARP_WIFI);
BRPORT_ATTR_FLAG(multicast_flood, BR_MCAST_FLOOD);
BRPORT_ATTR_FLAG(broadcast_flood, BR_BCAST_FLOOD);
+BRPORT_ATTR_FLAG(isolate_mode, BR_ISOLATE_MODE);
#ifdef CONFIG_BRIDGE_IGMP_SNOOPING
static ssize_t show_multicast_router(struct net_bridge_port *p, char *buf)
@@ -223,6 +224,7 @@ static const struct brport_attribute *br
&brport_attr_proxyarp_wifi,
&brport_attr_multicast_flood,
&brport_attr_broadcast_flood,
+ &brport_attr_isolate_mode,
NULL
};

24
target/linux/generic/hack-4.14/651-wireless_mesh_header.patch Normal file
View file

@ -0,0 +1,24 @@
From 6d3bc769657b0ee7c7506dad9911111c4226a7ea Mon Sep 17 00:00:00 2001
From: Imre Kaloz <kaloz@openwrt.org>
Date: Fri, 7 Jul 2017 17:21:05 +0200
Subject: mac80211: increase wireless mesh header size
lede-commit 3d4466cfd8f75f717efdb1f96fdde3c70d865fc1
Signed-off-by: Imre Kaloz <kaloz@openwrt.org>
---
include/linux/netdevice.h | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -138,8 +138,8 @@ static inline bool dev_xmit_complete(int
#if defined(CONFIG_HYPERV_NET)
# define LL_MAX_HEADER 128
-#elif defined(CONFIG_WLAN) || IS_ENABLED(CONFIG_AX25)
-# if defined(CONFIG_MAC80211_MESH)
+#elif defined(CONFIG_WLAN) || IS_ENABLED(CONFIG_AX25) || 1
+# if defined(CONFIG_MAC80211_MESH) || 1
# define LL_MAX_HEADER 128
# else
# define LL_MAX_HEADER 96

27
target/linux/generic/hack-4.14/660-fq_codel_defaults.patch Normal file
View file

@ -0,0 +1,27 @@
From a6ccb238939b25851474a279b20367fd24a0e816 Mon Sep 17 00:00:00 2001
From: Felix Fietkau <nbd@nbd.name>
Date: Fri, 7 Jul 2017 17:21:53 +0200
Subject: hack: net: fq_codel: tune defaults for small devices
Assume that x86_64 devices always have a big memory and do not need this
optimization compared to devices with only 32 MB or 64 MB RAM.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
net/sched/sch_fq_codel.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
--- a/net/sched/sch_fq_codel.c
+++ b/net/sched/sch_fq_codel.c
@@ -465,7 +465,11 @@ static int fq_codel_init(struct Qdisc *s
sch->limit = 10*1024;
q->flows_cnt = 1024;
+#ifdef CONFIG_X86_64
q->memory_limit = 32 << 20; /* 32 MBytes */
+#else
+ q->memory_limit = 4 << 20; /* 4 MBytes */
+#endif
q->drop_batch_size = 64;
q->quantum = psched_mtu(qdisc_dev(sch));
INIT_LIST_HEAD(&q->new_flows);

94
target/linux/generic/hack-4.14/661-use_fq_codel_by_default.patch Normal file
View file

@ -0,0 +1,94 @@
From 1d418f7e88035ed7a94073f6354246c66e9193e9 Mon Sep 17 00:00:00 2001
From: Felix Fietkau <nbd@nbd.name>
Date: Fri, 7 Jul 2017 17:22:58 +0200
Subject: fq_codel: switch default qdisc from pfifo_fast to fq_codel and remove pfifo_fast
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
include/net/sch_generic.h | 3 ++-
net/sched/Kconfig | 3 ++-
net/sched/sch_api.c | 2 +-
net/sched/sch_fq_codel.c | 3 ++-
net/sched/sch_generic.c | 4 ++--
5 files changed, 9 insertions(+), 6 deletions(-)
--- a/include/net/sch_generic.h
+++ b/include/net/sch_generic.h
@@ -369,12 +369,13 @@ extern struct Qdisc_ops noop_qdisc_ops;
extern struct Qdisc_ops pfifo_fast_ops;
extern struct Qdisc_ops mq_qdisc_ops;
extern struct Qdisc_ops noqueue_qdisc_ops;
+extern struct Qdisc_ops fq_codel_qdisc_ops;
extern const struct Qdisc_ops *default_qdisc_ops;
static inline const struct Qdisc_ops *
get_default_qdisc_ops(const struct net_device *dev, int ntx)
{
return ntx < dev->real_num_tx_queues ?
- default_qdisc_ops : &pfifo_fast_ops;
+ default_qdisc_ops : &fq_codel_qdisc_ops;
}
struct Qdisc_class_common {
--- a/net/sched/Kconfig
+++ b/net/sched/Kconfig
@@ -3,8 +3,9 @@
#
menuconfig NET_SCHED
- bool "QoS and/or fair queueing"
+ def_bool y
select NET_SCH_FIFO
+ select NET_SCH_FQ_CODEL
---help---
When the kernel has several packets to send out over a network
device, it has to decide which ones to send first, which ones to
--- a/net/sched/sch_api.c
+++ b/net/sched/sch_api.c
@@ -2014,7 +2014,7 @@ static int __init pktsched_init(void)
return err;
}
- register_qdisc(&pfifo_fast_ops);
+ register_qdisc(&fq_codel_qdisc_ops);
register_qdisc(&pfifo_qdisc_ops);
register_qdisc(&bfifo_qdisc_ops);
register_qdisc(&pfifo_head_drop_qdisc_ops);
--- a/net/sched/sch_fq_codel.c
+++ b/net/sched/sch_fq_codel.c
@@ -694,7 +694,7 @@ static const struct Qdisc_class_ops fq_c
.walk = fq_codel_walk,
};
-static struct Qdisc_ops fq_codel_qdisc_ops __read_mostly = {
+struct Qdisc_ops fq_codel_qdisc_ops __read_mostly = {
.cl_ops = &fq_codel_class_ops,
.id = "fq_codel",
.priv_size = sizeof(struct fq_codel_sched_data),
@@ -709,6 +709,7 @@ static struct Qdisc_ops fq_codel_qdisc_o
.dump_stats = fq_codel_dump_stats,
.owner = THIS_MODULE,
};
+EXPORT_SYMBOL(fq_codel_qdisc_ops);
static int __init fq_codel_module_init(void)
{
--- a/net/sched/sch_generic.c
+++ b/net/sched/sch_generic.c
@@ -32,7 +32,7 @@
#include <trace/events/qdisc.h>
/* Qdisc to use by default */
-const struct Qdisc_ops *default_qdisc_ops = &pfifo_fast_ops;
+const struct Qdisc_ops *default_qdisc_ops = &fq_codel_qdisc_ops;
EXPORT_SYMBOL(default_qdisc_ops);
/* Main transmission queue. */
@@ -764,7 +764,7 @@ static void attach_one_default_qdisc(str
void *_unused)
{
struct Qdisc *qdisc;
- const struct Qdisc_ops *ops = default_qdisc_ops;
+ const struct Qdisc_ops *ops = &fq_codel_qdisc_ops;
if (dev->priv_flags & IFF_NO_QUEUE)
ops = &noqueue_qdisc_ops;

159
target/linux/generic/hack-4.14/662-remove_pfifo_fast.patch Normal file
View file

@ -0,0 +1,159 @@
From b531d492d5ef1cf9dba0f4888eb5fd8624a6d762 Mon Sep 17 00:00:00 2001
From: Felix Fietkau <nbd@nbd.name>
Date: Fri, 7 Jul 2017 17:23:42 +0200
Subject: net: sched: switch default qdisc from pfifo_fast to fq_codel and remove pfifo_fast
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
net/sched/sch_generic.c | 140 ------------------------------------------------
1 file changed, 140 deletions(-)
--- a/net/sched/sch_generic.c
+++ b/net/sched/sch_generic.c
@@ -453,146 +453,6 @@ struct Qdisc_ops noqueue_qdisc_ops __rea
.owner = THIS_MODULE,
};
-static const u8 prio2band[TC_PRIO_MAX + 1] = {
- 1, 2, 2, 2, 1, 2, 0, 0 , 1, 1, 1, 1, 1, 1, 1, 1
-};
-
-/* 3-band FIFO queue: old style, but should be a bit faster than
- generic prio+fifo combination.
- */
-
-#define PFIFO_FAST_BANDS 3
-
-/*
- * Private data for a pfifo_fast scheduler containing:
- * - queues for the three band
- * - bitmap indicating which of the bands contain skbs
- */
-struct pfifo_fast_priv {
- u32 bitmap;
- struct qdisc_skb_head q[PFIFO_FAST_BANDS];
-};
-
-/*
- * Convert a bitmap to the first band number where an skb is queued, where:
- * bitmap=0 means there are no skbs on any band.
- * bitmap=1 means there is an skb on band 0.
- * bitmap=7 means there are skbs on all 3 bands, etc.
- */
-static const int bitmap2band[] = {-1, 0, 1, 0, 2, 0, 1, 0};
-
-static inline struct qdisc_skb_head *band2list(struct pfifo_fast_priv *priv,
- int band)
-{
- return priv->q + band;
-}
-
-static int pfifo_fast_enqueue(struct sk_buff *skb, struct Qdisc *qdisc,
- struct sk_buff **to_free)
-{
- if (qdisc->q.qlen < qdisc_dev(qdisc)->tx_queue_len) {
- int band = prio2band[skb->priority & TC_PRIO_MAX];
- struct pfifo_fast_priv *priv = qdisc_priv(qdisc);
- struct qdisc_skb_head *list = band2list(priv, band);
-
- priv->bitmap |= (1 << band);
- qdisc->q.qlen++;
- return __qdisc_enqueue_tail(skb, qdisc, list);
- }
-
- return qdisc_drop(skb, qdisc, to_free);
-}
-
-static struct sk_buff *pfifo_fast_dequeue(struct Qdisc *qdisc)
-{
- struct pfifo_fast_priv *priv = qdisc_priv(qdisc);
- int band = bitmap2band[priv->bitmap];
-
- if (likely(band >= 0)) {
- struct qdisc_skb_head *qh = band2list(priv, band);
- struct sk_buff *skb = __qdisc_dequeue_head(qh);
-
- if (likely(skb != NULL)) {
- qdisc_qstats_backlog_dec(qdisc, skb);
- qdisc_bstats_update(qdisc, skb);
- }
-
- qdisc->q.qlen--;
- if (qh->qlen == 0)
- priv->bitmap &= ~(1 << band);
-
- return skb;
- }
-
- return NULL;
-}
-
-static struct sk_buff *pfifo_fast_peek(struct Qdisc *qdisc)
-{
- struct pfifo_fast_priv *priv = qdisc_priv(qdisc);
- int band = bitmap2band[priv->bitmap];
-
- if (band >= 0) {
- struct qdisc_skb_head *qh = band2list(priv, band);
-
- return qh->head;
- }
-
- return NULL;
-}
-
-static void pfifo_fast_reset(struct Qdisc *qdisc)
-{
- int prio;
- struct pfifo_fast_priv *priv = qdisc_priv(qdisc);
-
- for (prio = 0; prio < PFIFO_FAST_BANDS; prio++)
- __qdisc_reset_queue(band2list(priv, prio));
-
- priv->bitmap = 0;
- qdisc->qstats.backlog = 0;
- qdisc->q.qlen = 0;
-}
-
-static int pfifo_fast_dump(struct Qdisc *qdisc, struct sk_buff *skb)
-{
- struct tc_prio_qopt opt = { .bands = PFIFO_FAST_BANDS };
-
- memcpy(&opt.priomap, prio2band, TC_PRIO_MAX + 1);
- if (nla_put(skb, TCA_OPTIONS, sizeof(opt), &opt))
- goto nla_put_failure;
- return skb->len;
-
-nla_put_failure:
- return -1;
-}
-
-static int pfifo_fast_init(struct Qdisc *qdisc, struct nlattr *opt)
-{
- int prio;
- struct pfifo_fast_priv *priv = qdisc_priv(qdisc);
-
- for (prio = 0; prio < PFIFO_FAST_BANDS; prio++)
- qdisc_skb_head_init(band2list(priv, prio));
-
- /* Can by-pass the queue discipline */
- qdisc->flags |= TCQ_F_CAN_BYPASS;
- return 0;
-}
-
-struct Qdisc_ops pfifo_fast_ops __read_mostly = {
- .id = "pfifo_fast",
- .priv_size = sizeof(struct pfifo_fast_priv),
- .enqueue = pfifo_fast_enqueue,
- .dequeue = pfifo_fast_dequeue,
- .peek = pfifo_fast_peek,
- .init = pfifo_fast_init,
- .reset = pfifo_fast_reset,
- .dump = pfifo_fast_dump,
- .owner = THIS_MODULE,
-};
-EXPORT_SYMBOL(pfifo_fast_ops);
-
static struct lock_class_key qdisc_tx_busylock;
static struct lock_class_key qdisc_running_key;

128
target/linux/generic/hack-4.14/700-swconfig_switch_drivers.patch Normal file
View file

@ -0,0 +1,128 @@
From 36e516290611e613aa92996cb4339561452695b4 Mon Sep 17 00:00:00 2001
From: Felix Fietkau <nbd@nbd.name>
Date: Fri, 7 Jul 2017 17:24:23 +0200
Subject: net: swconfig: adds openwrt switch layer
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
drivers/net/phy/Kconfig | 83 +++++++++++++++++++++++++++++++++++++++++++++++
drivers/net/phy/Makefile | 15 +++++++++
include/uapi/linux/Kbuild | 1 +
3 files changed, 99 insertions(+)
--- a/drivers/net/phy/Kconfig
+++ b/drivers/net/phy/Kconfig
@@ -198,6 +198,89 @@ config LED_TRIGGER_PHY
<Speed in megabits>Mbps or <Speed in gigabits>Gbps
+comment "Switch configuration API + drivers"
+
+config SWCONFIG
+ tristate "Switch configuration API"
+ ---help---
+ Switch configuration API using netlink. This allows
+ you to configure the VLAN features of certain switches.
+
+config SWCONFIG_LEDS
+ bool "Switch LED trigger support"
+ depends on (SWCONFIG && LEDS_TRIGGERS)
+
+config ADM6996_PHY
+ tristate "Driver for ADM6996 switches"
+ select SWCONFIG
+ ---help---
+ Currently supports the ADM6996FC and ADM6996M switches.
+ Support for FC is very limited.
+
+config AR8216_PHY
+ tristate "Driver for Atheros AR8216 switches"
+ select ETHERNET_PACKET_MANGLE
+ select SWCONFIG
+
+config AR8216_PHY_LEDS
+ bool "Atheros AR8216 switch LED support"
+ depends on (AR8216_PHY && LEDS_CLASS)
+
+source "drivers/net/phy/b53/Kconfig"
+
+config IP17XX_PHY
+ tristate "Driver for IC+ IP17xx switches"
+ select SWCONFIG
+
+config MVSWITCH_PHY
+ tristate "Driver for Marvell 88E6060 switches"
+ select ETHERNET_PACKET_MANGLE
+
+config MVSW61XX_PHY
+ tristate "Driver for Marvell 88E6171/6172 switches"
+ select SWCONFIG
+
+config PSB6970_PHY
+ tristate "Lantiq XWAY Tantos (PSB6970) Ethernet switch"
+ select SWCONFIG
+ select ETHERNET_PACKET_MANGLE
+
+config RTL8306_PHY
+ tristate "Driver for Realtek RTL8306S switches"
+ select SWCONFIG
+
+config RTL8366_SMI
+ tristate "Driver for the RTL8366 SMI interface"
+ depends on GPIOLIB
+ ---help---
+ This module implements the SMI interface protocol which is used
+ by some RTL8366 ethernet switch devices via the generic GPIO API.
+
+if RTL8366_SMI
+
+config RTL8366_SMI_DEBUG_FS
+ bool "RTL8366 SMI interface debugfs support"
+ depends on DEBUG_FS
+ default n
+
+config RTL8366S_PHY
+ tristate "Driver for the Realtek RTL8366S switch"
+ select SWCONFIG
+
+config RTL8366RB_PHY
+ tristate "Driver for the Realtek RTL8366RB switch"
+ select SWCONFIG
+
+config RTL8367_PHY
+ tristate "Driver for the Realtek RTL8367R/M switches"
+ select SWCONFIG
+
+config RTL8367B_PHY
+ tristate "Driver fot the Realtek RTL8367R-VB switch"
+ select SWCONFIG
+
+endif # RTL8366_SMI
+
comment "MII PHY device drivers"
config SFP
--- a/drivers/net/phy/Makefile
+++ b/drivers/net/phy/Makefile
@@ -22,6 +22,21 @@ libphy-$(CONFIG_LED_TRIGGER_PHY) += phy_
obj-$(CONFIG_PHYLINK) += phylink.o
obj-$(CONFIG_PHYLIB) += libphy.o
+obj-$(CONFIG_SWCONFIG) += swconfig.o
+obj-$(CONFIG_ADM6996_PHY) += adm6996.o
+obj-$(CONFIG_AR8216_PHY) += ar8216.o ar8327.o
+obj-$(CONFIG_SWCONFIG_B53) += b53/
+obj-$(CONFIG_IP17XX_PHY) += ip17xx.o
+obj-$(CONFIG_MVSWITCH_PHY) += mvswitch.o
+obj-$(CONFIG_MVSW61XX_PHY) += mvsw61xx.o
+obj-$(CONFIG_PSB6970_PHY) += psb6970.o
+obj-$(CONFIG_RTL8306_PHY) += rtl8306.o
+obj-$(CONFIG_RTL8366_SMI) += rtl8366_smi.o
+obj-$(CONFIG_RTL8366S_PHY) += rtl8366s.o
+obj-$(CONFIG_RTL8366RB_PHY) += rtl8366rb.o
+obj-$(CONFIG_RTL8367_PHY) += rtl8367.o
+obj-$(CONFIG_RTL8367B_PHY) += rtl8367b.o
+
obj-$(CONFIG_MDIO_BCM_IPROC) += mdio-bcm-iproc.o
obj-$(CONFIG_MDIO_BCM_UNIMAC) += mdio-bcm-unimac.o
obj-$(CONFIG_MDIO_BITBANG) += mdio-bitbang.o

27
target/linux/generic/hack-4.14/702-phy_add_aneg_done_function.patch Normal file
View file

@ -0,0 +1,27 @@
--- a/include/linux/phy.h
+++ b/include/linux/phy.h
@@ -547,6 +547,12 @@ struct phy_driver {
/* Determines the negotiated speed and duplex */
int (*read_status)(struct phy_device *phydev);
+ /*
+ * Update the value in phydev->link to reflect the
+ * current link value
+ */
+ int (*update_link)(struct phy_device *phydev);
+
/* Clears any pending interrupts */
int (*ack_interrupt)(struct phy_device *phydev);
--- a/drivers/net/phy/phy_device.c
+++ b/drivers/net/phy/phy_device.c
@@ -1437,6 +1437,9 @@ int genphy_update_link(struct phy_device
{
int status;
+ if (phydev->drv && phydev->drv->update_link)
+ return phydev->drv->update_link(phydev);
+
/* Do a fake read */
status = phy_read(phydev, MII_BMSR);
if (status < 0)

176
target/linux/generic/hack-4.14/721-phy_packets.patch Normal file
View file

@ -0,0 +1,176 @@
From ffe387740bbe88dd88bbe04d6375902708003d6e Mon Sep 17 00:00:00 2001
From: Felix Fietkau <nbd@nbd.name>
Date: Fri, 7 Jul 2017 17:25:00 +0200
Subject: net: add packet mangeling patch
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
include/linux/netdevice.h | 11 +++++++++++
include/linux/skbuff.h | 14 ++++----------
net/Kconfig | 6 ++++++
net/core/dev.c | 18 ++++++++++++++----
net/core/skbuff.c | 17 +++++++++++++++++
net/ethernet/eth.c | 6 ++++++
6 files changed, 58 insertions(+), 14 deletions(-)
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -1386,6 +1386,7 @@ enum netdev_priv_flags {
IFF_RXFH_CONFIGURED = 1<<25,
IFF_PHONY_HEADROOM = 1<<26,
IFF_MACSEC = 1<<27,
+ IFF_NO_IP_ALIGN = 1<<28,
};
#define IFF_802_1Q_VLAN IFF_802_1Q_VLAN
@@ -1415,6 +1416,7 @@ enum netdev_priv_flags {
#define IFF_TEAM IFF_TEAM
#define IFF_RXFH_CONFIGURED IFF_RXFH_CONFIGURED
#define IFF_MACSEC IFF_MACSEC
+#define IFF_NO_IP_ALIGN IFF_NO_IP_ALIGN
/**
* struct net_device - The DEVICE structure.
@@ -1701,6 +1703,11 @@ struct net_device {
const struct xfrmdev_ops *xfrmdev_ops;
#endif
+#ifdef CONFIG_ETHERNET_PACKET_MANGLE
+ void (*eth_mangle_rx)(struct net_device *dev, struct sk_buff *skb);
+ struct sk_buff *(*eth_mangle_tx)(struct net_device *dev, struct sk_buff *skb);
+#endif
+
const struct header_ops *header_ops;
unsigned int flags;
@@ -1770,6 +1777,10 @@ struct net_device {
struct mpls_dev __rcu *mpls_ptr;
#endif
+#ifdef CONFIG_ETHERNET_PACKET_MANGLE
+ void *phy_ptr; /* PHY device specific data */
+#endif
+
/*
* Cache lines mostly used on receive path (including eth_type_trans())
*/
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -2492,6 +2492,10 @@ static inline int pskb_trim(struct sk_bu
return (len < skb->len) ? __pskb_trim(skb, len) : 0;
}
+extern struct sk_buff *__netdev_alloc_skb_ip_align(struct net_device *dev,
+ unsigned int length, gfp_t gfp);
+
+
/**
* pskb_trim_unique - remove end from a paged unique (not cloned) buffer
* @skb: buffer to alter
@@ -2622,16 +2626,6 @@ static inline struct sk_buff *dev_alloc_
}
-static inline struct sk_buff *__netdev_alloc_skb_ip_align(struct net_device *dev,
- unsigned int length, gfp_t gfp)
-{
- struct sk_buff *skb = __netdev_alloc_skb(dev, length + NET_IP_ALIGN, gfp);
-
- if (NET_IP_ALIGN && skb)
- skb_reserve(skb, NET_IP_ALIGN);
- return skb;
-}
-
static inline struct sk_buff *netdev_alloc_skb_ip_align(struct net_device *dev,
unsigned int length)
{
--- a/net/Kconfig
+++ b/net/Kconfig
@@ -25,6 +25,12 @@ menuconfig NET
if NET
+config ETHERNET_PACKET_MANGLE
+ bool
+ help
+ This option can be selected by phy drivers that need to mangle
+ packets going in or out of an ethernet device.
+
config WANT_COMPAT_NETLINK_MESSAGES
bool
help
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -2974,10 +2974,20 @@ static int xmit_one(struct sk_buff *skb,
if (!list_empty(&ptype_all) || !list_empty(&dev->ptype_all))
dev_queue_xmit_nit(skb, dev);
- len = skb->len;
- trace_net_dev_start_xmit(skb, dev);
- rc = netdev_start_xmit(skb, dev, txq, more);
- trace_net_dev_xmit(skb, rc, dev, len);
+#ifdef CONFIG_ETHERNET_PACKET_MANGLE
+ if (!dev->eth_mangle_tx ||
+ (skb = dev->eth_mangle_tx(dev, skb)) != NULL)
+#else
+ if (1)
+#endif
+ {
+ len = skb->len;
+ trace_net_dev_start_xmit(skb, dev);
+ rc = netdev_start_xmit(skb, dev, txq, more);
+ trace_net_dev_xmit(skb, rc, dev, len);
+ } else {
+ rc = NETDEV_TX_OK;
+ }
return rc;
}
--- a/net/core/skbuff.c
+++ b/net/core/skbuff.c
@@ -64,6 +64,7 @@
#include <linux/errqueue.h>
#include <linux/prefetch.h>
#include <linux/if_vlan.h>
+#include <linux/if.h>
#include <net/protocol.h>
#include <net/dst.h>
@@ -503,6 +504,22 @@ skb_fail:
}
EXPORT_SYMBOL(__napi_alloc_skb);
+struct sk_buff *__netdev_alloc_skb_ip_align(struct net_device *dev,
+ unsigned int length, gfp_t gfp)
+{
+ struct sk_buff *skb = __netdev_alloc_skb(dev, length + NET_IP_ALIGN, gfp);
+
+#ifdef CONFIG_ETHERNET_PACKET_MANGLE
+ if (dev && (dev->priv_flags & IFF_NO_IP_ALIGN))
+ return skb;
+#endif
+
+ if (NET_IP_ALIGN && skb)
+ skb_reserve(skb, NET_IP_ALIGN);
+ return skb;
+}
+EXPORT_SYMBOL(__netdev_alloc_skb_ip_align);
+
void skb_add_rx_frag(struct sk_buff *skb, int i, struct page *page, int off,
int size, unsigned int truesize)
{
--- a/net/ethernet/eth.c
+++ b/net/ethernet/eth.c
@@ -172,6 +172,12 @@ __be16 eth_type_trans(struct sk_buff *sk
const struct ethhdr *eth;
skb->dev = dev;
+
+#ifdef CONFIG_ETHERNET_PACKET_MANGLE
+ if (dev->eth_mangle_rx)
+ dev->eth_mangle_rx(dev, skb);
+#endif
+
skb_reset_mac_header(skb);
eth = (struct ethhdr *)skb->data;

98
target/linux/generic/hack-4.14/773-bgmac-add-srab-switch.patch Normal file
View file

@ -0,0 +1,98 @@
From 3cb240533ab787899dc7f17aa7d6c5b4810e2e58 Mon Sep 17 00:00:00 2001
From: Hauke Mehrtens <hauke@hauke-m.de>
Date: Fri, 7 Jul 2017 17:26:01 +0200
Subject: bcm53xx: bgmac: use srab switch driver
use the srab switch driver on these SoCs.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
---
drivers/net/ethernet/broadcom/bgmac-bcma.c | 1 +
drivers/net/ethernet/broadcom/bgmac.c | 24 ++++++++++++++++++++++++
drivers/net/ethernet/broadcom/bgmac.h | 4 ++++
3 files changed, 29 insertions(+)
--- a/drivers/net/ethernet/broadcom/bgmac-bcma.c
+++ b/drivers/net/ethernet/broadcom/bgmac-bcma.c
@@ -268,6 +268,7 @@ static int bgmac_probe(struct bcma_devic
bgmac->feature_flags |= BGMAC_FEAT_CLKCTLST;
bgmac->feature_flags |= BGMAC_FEAT_NO_RESET;
bgmac->feature_flags |= BGMAC_FEAT_FORCE_SPEED_2500;
+ bgmac->feature_flags |= BGMAC_FEAT_SRAB;
break;
default:
bgmac->feature_flags |= BGMAC_FEAT_CLKCTLST;
--- a/drivers/net/ethernet/broadcom/bgmac.c
+++ b/drivers/net/ethernet/broadcom/bgmac.c
@@ -12,6 +12,7 @@
#include <linux/bcma/bcma.h>
#include <linux/etherdevice.h>
#include <linux/interrupt.h>
+#include <linux/platform_data/b53.h>
#include <linux/bcm47xx_nvram.h>
#include <linux/phy.h>
#include <linux/phy_fixed.h>
@@ -1409,6 +1410,17 @@ static const struct ethtool_ops bgmac_et
.set_link_ksettings = phy_ethtool_set_link_ksettings,
};
+static struct b53_platform_data bgmac_b53_pdata = {
+};
+
+static struct platform_device bgmac_b53_dev = {
+ .name = "b53-srab-switch",
+ .id = -1,
+ .dev = {
+ .platform_data = &bgmac_b53_pdata,
+ },
+};
+
/**************************************************
* MII
**************************************************/
@@ -1538,6 +1550,14 @@ int bgmac_enet_probe(struct bgmac *bgmac
net_dev->hw_features = net_dev->features;
net_dev->vlan_features = net_dev->features;
+ if ((bgmac->feature_flags & BGMAC_FEAT_SRAB) && !bgmac_b53_pdata.regs) {
+ bgmac_b53_pdata.regs = ioremap_nocache(0x18007000, 0x1000);
+
+ err = platform_device_register(&bgmac_b53_dev);
+ if (!err)
+ bgmac->b53_device = &bgmac_b53_dev;
+ }
+
err = register_netdev(bgmac->net_dev);
if (err) {
dev_err(bgmac->dev, "Cannot register net device\n");
@@ -1560,6 +1580,10 @@ EXPORT_SYMBOL_GPL(bgmac_enet_probe);
void bgmac_enet_remove(struct bgmac *bgmac)
{
+ if (bgmac->b53_device)
+ platform_device_unregister(&bgmac_b53_dev);
+ bgmac->b53_device = NULL;
+
unregister_netdev(bgmac->net_dev);
phy_disconnect(bgmac->net_dev->phydev);
netif_napi_del(&bgmac->napi);
--- a/drivers/net/ethernet/broadcom/bgmac.h
+++ b/drivers/net/ethernet/broadcom/bgmac.h
@@ -427,6 +427,7 @@
#define BGMAC_FEAT_CC4_IF_SW_TYPE_RGMII BIT(18)
#define BGMAC_FEAT_CC7_IF_TYPE_RGMII BIT(19)
#define BGMAC_FEAT_IDM_MASK BIT(20)
+#define BGMAC_FEAT_SRAB BIT(21)
struct bgmac_slot_info {
union {
@@ -532,6 +533,9 @@ struct bgmac {
void (*cmn_maskset32)(struct bgmac *bgmac, u16 offset, u32 mask,
u32 set);
int (*phy_connect)(struct bgmac *bgmac);
+
+ /* platform device for associated switch */
+ struct platform_device *b53_device;
};
struct bgmac *bgmac_alloc(struct device *dev);

52
target/linux/generic/hack-4.14/835-misc-owl_loader.patch Normal file
View file

@ -0,0 +1,52 @@
From dd36f935973d91644449bd9749f6062a2bed821b Mon Sep 17 00:00:00 2001
From: Christian Lamparter <chunkeey@googlemail.com>
Date: Fri, 7 Jul 2017 17:26:46 +0200
Subject: misc: owl-loader for delayed Atheros ath9k fixup
Some devices (like the Cisco Meraki Z1 Cloud Managed Teleworker Gateway)
need to be able to initialize the PCIe wifi device. Normally, this is done
during the early stages of booting linux, because the necessary init code
is read from the memory mapped SPI and passed to pci_enable_ath9k_fixup.
However,this isn't possible for devices which have the init code for the
Atheros chip stored on NAND in an UBI volume. Hence, this module can be
used to initialze the chip when the user-space is ready to extract the
init code.
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Signed-off-by: Christian Lamparter <chunkeey@googlemail.com>
---
drivers/misc/Kconfig | 12 ++++++++++++
drivers/misc/Makefile | 1 +
2 files changed, 13 insertions(+)
--- a/drivers/misc/Kconfig
+++ b/drivers/misc/Kconfig
@@ -151,6 +151,18 @@ config SGI_IOC4
If you have an SGI Altix with an IOC4-based card say Y.
Otherwise say N.
+config OWL_LOADER
+ tristate "Owl loader for initializing Atheros PCI(e) Wifi chips"
+ depends on PCI
+ ---help---
+ This kernel module helps to initialize certain Qualcomm
+ Atheros' PCI(e) Wifi chips, which have the init data
+ (which contains the PCI device ID for example) stored
+ together with the calibration data in the file system.
+
+ This is necessary for devices like the Cisco Meraki Z1, say M.
+ Otherwise say N.
+
config TIFM_CORE
tristate "TI Flash Media interface support"
depends on PCI
--- a/drivers/misc/Makefile
+++ b/drivers/misc/Makefile
@@ -13,6 +13,7 @@ obj-$(CONFIG_ATMEL_TCLIB) += atmel_tclib
obj-$(CONFIG_DUMMY_IRQ) += dummy-irq.o
obj-$(CONFIG_ICS932S401) += ics932s401.o
obj-$(CONFIG_LKDTM) += lkdtm.o
+obj-$(CONFIG_OWL_LOADER) += owl-loader.o
obj-$(CONFIG_TIFM_CORE) += tifm_core.o
obj-$(CONFIG_TIFM_7XX1) += tifm_7xx1.o
obj-$(CONFIG_PHANTOM) += phantom.o

136
target/linux/generic/hack-4.14/901-debloat_sock_diag.patch Normal file
View file

@ -0,0 +1,136 @@
From 3b6115d6b57a263bdc8c9b1df273bd4a7955eead Mon Sep 17 00:00:00 2001
From: Felix Fietkau <nbd@nbd.name>
Date: Sat, 8 Jul 2017 08:16:31 +0200
Subject: debloat: add some debloat patches, strip down procfs and make O_DIRECT support optional, saves ~15K after lzma on MIPS
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
net/Kconfig | 3 +++
net/core/Makefile | 3 ++-
net/core/sock.c | 2 ++
net/ipv4/Kconfig | 1 +
net/netlink/Kconfig | 1 +
net/packet/Kconfig | 1 +
net/unix/Kconfig | 1 +
7 files changed, 11 insertions(+), 1 deletion(-)
--- a/net/Kconfig
+++ b/net/Kconfig
@@ -97,6 +97,9 @@ source "net/netlabel/Kconfig"
endif # if INET
+config SOCK_DIAG
+ bool
+
config NETWORK_SECMARK
bool "Security Marking"
help
--- a/net/core/Makefile
+++ b/net/core/Makefile
@@ -10,9 +10,10 @@ obj-$(CONFIG_SYSCTL) += sysctl_net_core.
obj-y += dev.o ethtool.o dev_addr_lists.o dst.o netevent.o \
neighbour.o rtnetlink.o utils.o link_watch.o filter.o \
- sock_diag.o dev_ioctl.o tso.o sock_reuseport.o \
+ dev_ioctl.o tso.o sock_reuseport.o \
fib_notifier.o
+obj-$(CONFIG_SOCK_DIAG) += sock_diag.o
obj-y += net-sysfs.o
obj-$(CONFIG_PROC_FS) += net-procfs.o
obj-$(CONFIG_NET_PKTGEN) += pktgen.o
--- a/net/core/sock.c
+++ b/net/core/sock.c
@@ -528,6 +528,18 @@ discard_and_relse:
}
EXPORT_SYMBOL(__sk_receive_skb);
+u64 sock_gen_cookie(struct sock *sk)
+{
+ while (1) {
+ u64 res = atomic64_read(&sk->sk_cookie);
+
+ if (res)
+ return res;
+ res = atomic64_inc_return(&sock_net(sk)->cookie_gen);
+ atomic64_cmpxchg(&sk->sk_cookie, 0, res);
+ }
+}
+
struct dst_entry *__sk_dst_check(struct sock *sk, u32 cookie)
{
struct dst_entry *dst = __sk_dst_get(sk);
@@ -1597,9 +1609,11 @@ void sk_destruct(struct sock *sk)
static void __sk_free(struct sock *sk)
{
+#ifdef CONFIG_SOCK_DIAG
if (unlikely(sock_diag_has_destroy_listeners(sk) && sk->sk_net_refcnt))
sock_diag_broadcast_destroy(sk);
else
+#endif
sk_destruct(sk);
}
--- a/net/core/sock_diag.c
+++ b/net/core/sock_diag.c
@@ -19,18 +19,6 @@ static int (*inet_rcv_compat)(struct sk_
static DEFINE_MUTEX(sock_diag_table_mutex);
static struct workqueue_struct *broadcast_wq;
-u64 sock_gen_cookie(struct sock *sk)
-{
- while (1) {
- u64 res = atomic64_read(&sk->sk_cookie);
-
- if (res)
- return res;
- res = atomic64_inc_return(&sock_net(sk)->cookie_gen);
- atomic64_cmpxchg(&sk->sk_cookie, 0, res);
- }
-}
-
int sock_diag_check_cookie(struct sock *sk, const __u32 *cookie)
{
u64 res;
--- a/net/ipv4/Kconfig
+++ b/net/ipv4/Kconfig
@@ -420,6 +420,7 @@ config INET_XFRM_MODE_BEET
config INET_DIAG
tristate "INET: socket monitoring interface"
+ select SOCK_DIAG
default y
---help---
Support for INET (TCP, DCCP, etc) socket monitoring interface used by
--- a/net/netlink/Kconfig
+++ b/net/netlink/Kconfig
@@ -4,6 +4,7 @@
config NETLINK_DIAG
tristate "NETLINK: socket monitoring interface"
+ select SOCK_DIAG
default n
---help---
Support for NETLINK socket monitoring interface used by the ss tool.
--- a/net/packet/Kconfig
+++ b/net/packet/Kconfig
@@ -18,6 +18,7 @@ config PACKET
config PACKET_DIAG
tristate "Packet: sockets monitoring interface"
depends on PACKET
+ select SOCK_DIAG
default n
---help---
Support for PF_PACKET sockets monitoring interface used by the ss tool.
--- a/net/unix/Kconfig
+++ b/net/unix/Kconfig
@@ -22,6 +22,7 @@ config UNIX
config UNIX_DIAG
tristate "UNIX: socket monitoring interface"
depends on UNIX
+ select SOCK_DIAG
default n
---help---
Support for UNIX socket monitoring interface used by the ss tool.

405
target/linux/generic/hack-4.14/902-debloat_proc.patch Normal file
View file

@ -0,0 +1,405 @@
From 9e3f1d0805b2d919904dd9a4ff0d956314cc3cba Mon Sep 17 00:00:00 2001
From: Felix Fietkau <nbd@nbd.name>
Date: Sat, 8 Jul 2017 08:20:09 +0200
Subject: debloat: procfs
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
fs/locks.c | 2 ++
fs/proc/Kconfig | 5 +++++
fs/proc/consoles.c | 3 +++
fs/proc/proc_tty.c | 11 ++++++++++-
include/net/snmp.h | 18 +++++++++++++++++-
ipc/msg.c | 3 +++
ipc/sem.c | 2 ++
ipc/shm.c | 2 ++
ipc/util.c | 3 +++
kernel/exec_domain.c | 2 ++
kernel/irq/proc.c | 9 +++++++++
kernel/time/timer_list.c | 2 ++
mm/vmalloc.c | 2 ++
mm/vmstat.c | 8 +++++---
net/8021q/vlanproc.c | 6 ++++++
net/core/net-procfs.c | 18 ++++++++++++------
net/core/sock.c | 2 ++
net/ipv4/fib_trie.c | 18 ++++++++++++------
net/ipv4/proc.c | 3 +++
net/ipv4/route.c | 3 +++
20 files changed, 105 insertions(+), 17 deletions(-)
--- a/fs/locks.c
+++ b/fs/locks.c
@@ -2805,6 +2805,8 @@ static const struct file_operations proc
static int __init proc_locks_init(void)
{
+ if (IS_ENABLED(CONFIG_PROC_STRIPPED))
+ return 0;
proc_create("locks", 0, NULL, &proc_locks_operations);
return 0;
}
--- a/fs/proc/Kconfig
+++ b/fs/proc/Kconfig
@@ -81,3 +81,8 @@ config PROC_CHILDREN
Say Y if you are running any user-space software which takes benefit from
this interface. For example, rkt is such a piece of software.
+
+config PROC_STRIPPED
+ default n
+ depends on EXPERT
+ bool "Strip non-essential /proc functionality to reduce code size"
--- a/fs/proc/consoles.c
+++ b/fs/proc/consoles.c
@@ -106,6 +106,9 @@ static const struct file_operations proc
static int __init proc_consoles_init(void)
{
+ if (IS_ENABLED(CONFIG_PROC_STRIPPED))
+ return 0;
+
proc_create("consoles", 0, NULL, &proc_consoles_operations);
return 0;
}
--- a/fs/proc/proc_tty.c
+++ b/fs/proc/proc_tty.c
@@ -144,7 +144,10 @@ static const struct file_operations proc
void proc_tty_register_driver(struct tty_driver *driver)
{
struct proc_dir_entry *ent;
-
+
+ if (IS_ENABLED(CONFIG_PROC_STRIPPED))
+ return;
+
if (!driver->driver_name || driver->proc_entry ||
!driver->ops->proc_fops)
return;
@@ -161,6 +164,9 @@ void proc_tty_unregister_driver(struct t
{
struct proc_dir_entry *ent;
+ if (IS_ENABLED(CONFIG_PROC_STRIPPED))
+ return;
+
ent = driver->proc_entry;
if (!ent)
return;
@@ -175,6 +181,9 @@ void proc_tty_unregister_driver(struct t
*/
void __init proc_tty_init(void)
{
+ if (IS_ENABLED(CONFIG_PROC_STRIPPED))
+ return;
+
if (!proc_mkdir("tty", NULL))
return;
proc_mkdir("tty/ldisc", NULL); /* Preserved: it's userspace visible */
--- a/include/net/snmp.h
+++ b/include/net/snmp.h
@@ -123,6 +123,21 @@ struct linux_xfrm_mib {
#define DECLARE_SNMP_STAT(type, name) \
extern __typeof__(type) __percpu *name
+#ifdef CONFIG_PROC_STRIPPED
+#define __SNMP_STATS_DUMMY(mib) \
+ do { (void) mib->mibs[0]; } while(0)
+
+#define __SNMP_INC_STATS(mib, field) __SNMP_STATS_DUMMY(mib)
+#define SNMP_INC_STATS_ATOMIC_LONG(mib, field) __SNMP_STATS_DUMMY(mib)
+#define SNMP_INC_STATS(mib, field) __SNMP_STATS_DUMMY(mib)
+#define SNMP_DEC_STATS(mib, field) __SNMP_STATS_DUMMY(mib)
+#define __SNMP_ADD_STATS(mib, field, addend) __SNMP_STATS_DUMMY(mib)
+#define SNMP_ADD_STATS(mib, field, addend) __SNMP_STATS_DUMMY(mib)
+#define SNMP_UPD_PO_STATS(mib, basefield, addend) __SNMP_STATS_DUMMY(mib)
+#define __SNMP_UPD_PO_STATS(mib, basefield, addend) __SNMP_STATS_DUMMY(mib)
+
+#else
+
#define __SNMP_INC_STATS(mib, field) \
__this_cpu_inc(mib->mibs[field])
@@ -153,8 +168,9 @@ struct linux_xfrm_mib {
__this_cpu_add(ptr[basefield##OCTETS], addend); \
} while (0)
+#endif
-#if BITS_PER_LONG==32
+#if (BITS_PER_LONG==32) && !defined(CONFIG_PROC_STRIPPED)
#define __SNMP_ADD_STATS64(mib, field, addend) \
do { \
--- a/ipc/msg.c
+++ b/ipc/msg.c
@@ -1208,6 +1208,9 @@ int __init msg_init(void)
{
const int err = msg_init_ns(&init_ipc_ns);
+ if (IS_ENABLED(CONFIG_PROC_STRIPPED))
+ return err;
+
ipc_init_proc_interface("sysvipc/msg",
" key msqid perms cbytes qnum lspid lrpid uid gid cuid cgid stime rtime ctime\n",
IPC_MSG_IDS, sysvipc_msg_proc_show);
--- a/ipc/sem.c
+++ b/ipc/sem.c
@@ -207,6 +207,8 @@ int __init sem_init(void)
{
const int err = sem_init_ns(&init_ipc_ns);
+ if (IS_ENABLED(CONFIG_PROC_STRIPPED))
+ return err;
ipc_init_proc_interface("sysvipc/sem",
" key semid perms nsems uid gid cuid cgid otime ctime\n",
IPC_SEM_IDS, sysvipc_sem_proc_show);
--- a/ipc/shm.c
+++ b/ipc/shm.c
@@ -122,6 +122,8 @@ pure_initcall(ipc_ns_init);
void __init shm_init(void)
{
+ if (IS_ENABLED(CONFIG_PROC_STRIPPED))
+ return;
ipc_init_proc_interface("sysvipc/shm",
#if BITS_PER_LONG <= 32
" key shmid perms size cpid lpid nattch uid gid cuid cgid atime dtime ctime rss swap\n",
--- a/ipc/util.c
+++ b/ipc/util.c
@@ -141,6 +141,9 @@ void __init ipc_init_proc_interface(cons
struct proc_dir_entry *pde;
struct ipc_proc_iface *iface;
+ if (IS_ENABLED(CONFIG_PROC_STRIPPED))
+ return;
+
iface = kmalloc(sizeof(*iface), GFP_KERNEL);
if (!iface)
return;
--- a/kernel/exec_domain.c
+++ b/kernel/exec_domain.c
@@ -42,6 +42,8 @@ static const struct file_operations exec
static int __init proc_execdomains_init(void)
{
+ if (IS_ENABLED(CONFIG_PROC_STRIPPED))
+ return 0;
proc_create("execdomains", 0, NULL, &execdomains_proc_fops);
return 0;
}
--- a/kernel/irq/proc.c
+++ b/kernel/irq/proc.c
@@ -396,6 +396,9 @@ void register_irq_proc(unsigned int irq,
void __maybe_unused *irqp = (void *)(unsigned long) irq;
char name [MAX_NAMELEN];
+ if (IS_ENABLED(CONFIG_PROC_STRIPPED) && !IS_ENABLED(CONFIG_SMP))
+ return;
+
if (!root_irq_dir || (desc->irq_data.chip == &no_irq_chip))
return;
@@ -449,6 +452,9 @@ void unregister_irq_proc(unsigned int ir
{
char name [MAX_NAMELEN];
+ if (IS_ENABLED(CONFIG_PROC_STRIPPED) && !IS_ENABLED(CONFIG_SMP))
+ return;
+
if (!root_irq_dir || !desc->dir)
return;
#ifdef CONFIG_SMP
@@ -487,6 +493,9 @@ void init_irq_proc(void)
unsigned int irq;
struct irq_desc *desc;
+ if (IS_ENABLED(CONFIG_PROC_STRIPPED) && !IS_ENABLED(CONFIG_SMP))
+ return;
+
/* create /proc/irq */
root_irq_dir = proc_mkdir("irq", NULL);
if (!root_irq_dir)
--- a/kernel/time/timer_list.c
+++ b/kernel/time/timer_list.c
@@ -389,6 +389,8 @@ static int __init init_timer_list_procfs
{
struct proc_dir_entry *pe;
+ if (IS_ENABLED(CONFIG_PROC_STRIPPED))
+ return 0;
pe = proc_create("timer_list", 0444, NULL, &timer_list_fops);
if (!pe)
return -ENOMEM;
--- a/mm/vmalloc.c
+++ b/mm/vmalloc.c
@@ -2765,6 +2765,8 @@ static const struct file_operations proc
static int __init proc_vmalloc_init(void)
{
+ if (IS_ENABLED(CONFIG_PROC_STRIPPED))
+ return 0;
proc_create("vmallocinfo", S_IRUSR, NULL, &proc_vmalloc_operations);
return 0;
}
--- a/mm/vmstat.c
+++ b/mm/vmstat.c
@@ -1944,10 +1944,12 @@ void __init init_mm_internals(void)
start_shepherd_timer();
#endif
#ifdef CONFIG_PROC_FS
- proc_create("buddyinfo", 0444, NULL, &buddyinfo_file_operations);
- proc_create("pagetypeinfo", 0444, NULL, &pagetypeinfo_file_operations);
+ if (!IS_ENABLED(CONFIG_PROC_STRIPPED)) {
+ proc_create("buddyinfo", 0444, NULL, &buddyinfo_file_operations);
+ proc_create("pagetypeinfo", 0444, NULL, &pagetypeinfo_file_operations);
+ proc_create("zoneinfo", 0444, NULL, &zoneinfo_file_operations);
+ }
proc_create("vmstat", 0444, NULL, &vmstat_file_operations);
- proc_create("zoneinfo", 0444, NULL, &zoneinfo_file_operations);
#endif
}
--- a/net/8021q/vlanproc.c
+++ b/net/8021q/vlanproc.c
@@ -127,6 +127,9 @@ void vlan_proc_cleanup(struct net *net)
{
struct vlan_net *vn = net_generic(net, vlan_net_id);
+ if (IS_ENABLED(CONFIG_PROC_STRIPPED))
+ return;
+
if (vn->proc_vlan_conf)
remove_proc_entry(name_conf, vn->proc_vlan_dir);
@@ -146,6 +149,9 @@ int __net_init vlan_proc_init(struct net
{
struct vlan_net *vn = net_generic(net, vlan_net_id);
+ if (IS_ENABLED(CONFIG_PROC_STRIPPED))
+ return 0;
+
vn->proc_vlan_dir = proc_net_mkdir(net, name_root, net->proc_net);
if (!vn->proc_vlan_dir)
goto err;
--- a/net/core/net-procfs.c
+++ b/net/core/net-procfs.c
@@ -320,10 +320,12 @@ static int __net_init dev_proc_net_init(
if (!proc_create("dev", S_IRUGO, net->proc_net, &dev_seq_fops))
goto out;
- if (!proc_create("softnet_stat", S_IRUGO, net->proc_net,
+ if (!IS_ENABLED(CONFIG_PROC_STRIPPED) &&
+ !proc_create("softnet_stat", S_IRUGO, net->proc_net,
&softnet_seq_fops))
goto out_dev;
- if (!proc_create("ptype", S_IRUGO, net->proc_net, &ptype_seq_fops))
+ if (!IS_ENABLED(CONFIG_PROC_STRIPPED) &&
+ !proc_create("ptype", S_IRUGO, net->proc_net, &ptype_seq_fops))
goto out_softnet;
if (wext_proc_init(net))
@@ -332,9 +334,11 @@ static int __net_init dev_proc_net_init(
out:
return rc;
out_ptype:
- remove_proc_entry("ptype", net->proc_net);
+ if (!IS_ENABLED(CONFIG_PROC_STRIPPED))
+ remove_proc_entry("ptype", net->proc_net);
out_softnet:
- remove_proc_entry("softnet_stat", net->proc_net);
+ if (!IS_ENABLED(CONFIG_PROC_STRIPPED))
+ remove_proc_entry("softnet_stat", net->proc_net);
out_dev:
remove_proc_entry("dev", net->proc_net);
goto out;
@@ -344,8 +348,10 @@ static void __net_exit dev_proc_net_exit
{
wext_proc_exit(net);
- remove_proc_entry("ptype", net->proc_net);
- remove_proc_entry("softnet_stat", net->proc_net);
+ if (!IS_ENABLED(CONFIG_PROC_STRIPPED)) {
+ remove_proc_entry("ptype", net->proc_net);
+ remove_proc_entry("softnet_stat", net->proc_net);
+ }
remove_proc_entry("dev", net->proc_net);
}
--- a/net/core/sock.c
+++ b/net/core/sock.c
@@ -3383,6 +3383,8 @@ static __net_initdata struct pernet_oper
static int __init proto_init(void)
{
+ if (IS_ENABLED(CONFIG_PROC_STRIPPED))
+ return 0;
return register_pernet_subsys(&proto_net_ops);
}
--- a/net/ipv4/fib_trie.c
+++ b/net/ipv4/fib_trie.c
@@ -2731,10 +2731,12 @@ static const struct file_operations fib_
int __net_init fib_proc_init(struct net *net)
{
- if (!proc_create("fib_trie", S_IRUGO, net->proc_net, &fib_trie_fops))
+ if (!IS_ENABLED(CONFIG_PROC_STRIPPED) &&
+ !proc_create("fib_trie", S_IRUGO, net->proc_net, &fib_trie_fops))
goto out1;
- if (!proc_create("fib_triestat", S_IRUGO, net->proc_net,
+ if (!IS_ENABLED(CONFIG_PROC_STRIPPED) &&
+ !proc_create("fib_triestat", S_IRUGO, net->proc_net,
&fib_triestat_fops))
goto out2;
@@ -2744,17 +2746,21 @@ int __net_init fib_proc_init(struct net
return 0;
out3:
- remove_proc_entry("fib_triestat", net->proc_net);
+ if (!IS_ENABLED(CONFIG_PROC_STRIPPED))
+ remove_proc_entry("fib_triestat", net->proc_net);
out2:
- remove_proc_entry("fib_trie", net->proc_net);
+ if (!IS_ENABLED(CONFIG_PROC_STRIPPED))
+ remove_proc_entry("fib_trie", net->proc_net);
out1:
return -ENOMEM;
}
void __net_exit fib_proc_exit(struct net *net)
{
- remove_proc_entry("fib_trie", net->proc_net);
- remove_proc_entry("fib_triestat", net->proc_net);
+ if (!IS_ENABLED(CONFIG_PROC_STRIPPED)) {
+ remove_proc_entry("fib_trie", net->proc_net);
+ remove_proc_entry("fib_triestat", net->proc_net);
+ }
remove_proc_entry("route", net->proc_net);
}
--- a/net/ipv4/proc.c
+++ b/net/ipv4/proc.c
@@ -557,6 +557,9 @@ static __net_initdata struct pernet_oper
int __init ip_misc_proc_init(void)
{
+ if (IS_ENABLED(CONFIG_PROC_STRIPPED))
+ return 0;
+
return register_pernet_subsys(&ip_proc_ops);
}
--- a/net/ipv4/route.c
+++ b/net/ipv4/route.c
@@ -424,6 +424,9 @@ static struct pernet_operations ip_rt_pr
static int __init ip_rt_proc_init(void)
{
+ if (IS_ENABLED(CONFIG_PROC_STRIPPED))
+ return 0;
+
return register_pernet_subsys(&ip_rt_proc_ops);
}

64
target/linux/generic/hack-4.14/904-debloat_dma_buf.patch Normal file
View file

@ -0,0 +1,64 @@
From e3692cb2fcd5ba1244512a0f43b8118f65f1c375 Mon Sep 17 00:00:00 2001
From: Felix Fietkau <nbd@nbd.name>
Date: Sat, 8 Jul 2017 08:20:43 +0200
Subject: debloat: dmabuf
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
drivers/base/Kconfig | 2 +-
drivers/dma-buf/Makefile | 10 +++++++---
drivers/dma-buf/dma-buf.c | 4 +++-
kernel/sched/core.c | 1 +
4 files changed, 12 insertions(+), 5 deletions(-)
--- a/drivers/base/Kconfig
+++ b/drivers/base/Kconfig
@@ -243,7 +243,7 @@ config SOC_BUS
source "drivers/base/regmap/Kconfig"
config DMA_SHARED_BUFFER
- bool
+ tristate
default n
select ANON_INODES
help
--- a/drivers/dma-buf/Makefile
+++ b/drivers/dma-buf/Makefile
@@ -1,3 +1,7 @@
-obj-y := dma-buf.o dma-fence.o dma-fence-array.o reservation.o seqno-fence.o
-obj-$(CONFIG_SYNC_FILE) += sync_file.o
-obj-$(CONFIG_SW_SYNC) += sw_sync.o sync_debug.o
+obj-$(CONFIG_DMA_SHARED_BUFFER) := dma-shared-buffer.o
+
+dma-buf-objs-y := dma-buf.o dma-fence.o dma-fence-array.o reservation.o seqno-fence.o
+dma-buf-objs-$(CONFIG_SYNC_FILE) += sync_file.o
+dma-buf-objs-$(CONFIG_SW_SYNC) += sw_sync.o sync_debug.o
+
+dma-shared-buffer-objs := $(dma-buf-objs-y)
--- a/drivers/dma-buf/dma-buf.c
+++ b/drivers/dma-buf/dma-buf.c
@@ -34,6 +34,7 @@
#include <linux/poll.h>
#include <linux/reservation.h>
#include <linux/mm.h>
+#include <linux/module.h>
#include <uapi/linux/dma-buf.h>
@@ -1205,4 +1206,5 @@ static void __exit dma_buf_deinit(void)
{
dma_buf_uninit_debugfs();
}
-__exitcall(dma_buf_deinit);
+module_exit(dma_buf_deinit);
+MODULE_LICENSE("GPL");
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -2146,6 +2146,7 @@ int wake_up_state(struct task_struct *p,
{
return try_to_wake_up(p, state, 0);
}
+EXPORT_SYMBOL_GPL(wake_up_state);
/*
* Perform scheduler related setup for a newly forked process p.

32
target/linux/generic/hack-4.14/910-kobject_uevent.patch Normal file
View file

@ -0,0 +1,32 @@
From 0d37e6edc09c99e683dd91ca0e83bbc0df8477b3 Mon Sep 17 00:00:00 2001
From: Felix Fietkau <nbd@nbd.name>
Date: Sun, 16 Jul 2017 16:56:10 +0200
Subject: lib: add uevent_next_seqnum()
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
include/linux/kobject.h | 5 +++++
lib/kobject_uevent.c | 37 +++++++++++++++++++++++++++++++++++++
2 files changed, 42 insertions(+)
--- a/lib/kobject_uevent.c
+++ b/lib/kobject_uevent.c
@@ -176,6 +176,18 @@ out:
return r;
}
+u64 uevent_next_seqnum(void)
+{
+ u64 seq;
+
+ mutex_lock(&uevent_sock_mutex);
+ seq = ++uevent_seqnum;
+ mutex_unlock(&uevent_sock_mutex);
+
+ return seq;
+}
+EXPORT_SYMBOL_GPL(uevent_next_seqnum);
+
/**
* kobject_synth_uevent - send synthetic uevent with arguments
*

76
target/linux/generic/hack-4.14/911-kobject_add_broadcast_uevent.patch Normal file
View file

@ -0,0 +1,76 @@
From 0d37e6edc09c99e683dd91ca0e83bbc0df8477b3 Mon Sep 17 00:00:00 2001
From: Felix Fietkau <nbd@nbd.name>
Date: Sun, 16 Jul 2017 16:56:10 +0200
Subject: lib: add uevent_next_seqnum()
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
include/linux/kobject.h | 5 +++++
lib/kobject_uevent.c | 37 +++++++++++++++++++++++++++++++++++++
2 files changed, 42 insertions(+)
--- a/include/linux/kobject.h
+++ b/include/linux/kobject.h
@@ -32,6 +32,8 @@
#define UEVENT_NUM_ENVP 32 /* number of env pointers */
#define UEVENT_BUFFER_SIZE 2048 /* buffer for the variables */
+struct sk_buff;
+
#ifdef CONFIG_UEVENT_HELPER
/* path to the userspace helper executed on an event */
extern char uevent_helper[];
@@ -224,4 +226,7 @@ int kobject_synth_uevent(struct kobject
__printf(2, 3)
int add_uevent_var(struct kobj_uevent_env *env, const char *format, ...);
+int broadcast_uevent(struct sk_buff *skb, __u32 pid, __u32 group,
+ gfp_t allocation);
+
#endif /* _KOBJECT_H_ */
--- a/lib/kobject_uevent.c
+++ b/lib/kobject_uevent.c
@@ -599,6 +599,43 @@ int add_uevent_var(struct kobj_uevent_en
EXPORT_SYMBOL_GPL(add_uevent_var);
#if defined(CONFIG_NET)
+int broadcast_uevent(struct sk_buff *skb, __u32 pid, __u32 group,
+ gfp_t allocation)
+{
+ struct uevent_sock *ue_sk;
+ int err = 0;
+
+ /* send netlink message */
+ mutex_lock(&uevent_sock_mutex);
+ list_for_each_entry(ue_sk, &uevent_sock_list, list) {
+ struct sock *uevent_sock = ue_sk->sk;
+ struct sk_buff *skb2;
+
+ skb2 = skb_clone(skb, allocation);
+ if (!skb2)
+ break;
+
+ err = netlink_broadcast(uevent_sock, skb2, pid, group,
+ allocation);
+ if (err)
+ break;
+ }
+ mutex_unlock(&uevent_sock_mutex);
+
+ kfree_skb(skb);
+ return err;
+}
+#else
+int broadcast_uevent(struct sk_buff *skb, __u32 pid, __u32 group,
+ gfp_t allocation)
+{
+ kfree_skb(skb);
+ return 0;
+}
+#endif
+EXPORT_SYMBOL_GPL(broadcast_uevent);
+
+#if defined(CONFIG_NET)
static int uevent_net_init(struct net *net)
{
struct uevent_sock *ue_sk;

40
target/linux/generic/hack-4.14/921-always-create-console-node-in-initramfs.patch Normal file
View file

@ -0,0 +1,40 @@
From 5d301596fdc72f6cb672f72eb3c66e7cddefb103 Mon Sep 17 00:00:00 2001
From: Felix Fietkau <nbd@nbd.name>
Date: Sat, 8 Jul 2017 08:26:02 +0200
Subject: initramfs: always create console node
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
scripts/gen_initramfs_list.sh | 14 ++++++++++++++
1 file changed, 14 insertions(+)
--- a/scripts/gen_initramfs_list.sh
+++ b/scripts/gen_initramfs_list.sh
@@ -59,6 +59,18 @@ default_initramfs() {
EOF
}
+list_openwrt_initramfs() {
+ :
+}
+
+openwrt_initramfs() {
+ # make sure that /dev/console exists
+ cat <<-EOF >> ${output}
+ dir /dev 0755 0 0
+ nod /dev/console 0600 0 0 c 5 1
+ EOF
+}
+
filetype() {
local argv1="$1"
@@ -180,6 +192,8 @@ dir_filelist() {
if [ "$(echo "${dirlist}" | wc -l)" -gt 1 ]; then
${dep_list}print_mtime "$1"
+ ${dep_list}openwrt_initramfs
+
echo "${dirlist}" | \
while read x; do
${dep_list}parse ${x}

338
target/linux/generic/hack-4.14/930-crashlog.patch Normal file
View file

@ -0,0 +1,338 @@
From 6b1ab74a9917012d0c559edc4ed299d9228ac89f Mon Sep 17 00:00:00 2001
From: Felix Fietkau <nbd@nbd.name>
Date: Sat, 8 Jul 2017 08:26:47 +0200
Subject: kernel: add the new 'crashlog' feature
this tries to store kernel oops/panic logs in a fixed location in RAM to
recover them available to user space using debugfs
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
include/linux/crashlog.h | 17 ++++
init/Kconfig | 4 +
kernel/Makefile | 1 +
kernel/crashlog.c | 213 +++++++++++++++++++++++++++++++++++++++++++++++
kernel/module.c | 3 +
mm/bootmem.c | 2 +
mm/memblock.c | 5 ++
7 files changed, 245 insertions(+)
create mode 100644 include/linux/crashlog.h
create mode 100644 kernel/crashlog.c
--- /dev/null
+++ b/include/linux/crashlog.h
@@ -0,0 +1,17 @@
+#ifndef __CRASHLOG_H
+#define __CRASHLOG_H
+
+#ifdef CONFIG_CRASHLOG
+void crashlog_init_bootmem(struct bootmem_data *bdata);
+void crashlog_init_memblock(phys_addr_t addr, phys_addr_t size);
+#else
+static inline void crashlog_init_bootmem(struct bootmem_data *bdata)
+{
+}
+
+static inline void crashlog_init_memblock(phys_addr_t addr, phys_addr_t size)
+{
+}
+#endif
+
+#endif
--- a/init/Kconfig
+++ b/init/Kconfig
@@ -1009,6 +1009,10 @@ config RELAY
If unsure, say N.
+config CRASHLOG
+ bool "Crash logging"
+ depends on (!NO_BOOTMEM || HAVE_MEMBLOCK)
+
config BLK_DEV_INITRD
bool "Initial RAM filesystem and RAM disk (initramfs/initrd) support"
depends on BROKEN || !FRV
--- a/kernel/Makefile
+++ b/kernel/Makefile
@@ -112,6 +112,7 @@ obj-$(CONFIG_CONTEXT_TRACKING) += contex
obj-$(CONFIG_TORTURE_TEST) += torture.o
obj-$(CONFIG_HAS_IOMEM) += memremap.o
+obj-$(CONFIG_CRASHLOG) += crashlog.o
$(obj)/configs.o: $(obj)/config_data.h
--- /dev/null
+++ b/kernel/crashlog.c
@@ -0,0 +1,213 @@
+/*
+ * Crash information logger
+ * Copyright (C) 2010 Felix Fietkau <nbd@nbd.name>
+ *
+ * Based on ramoops.c
+ * Copyright (C) 2010 Marco Stornelli <marco.stornelli@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA
+ * 02110-1301 USA
+ *
+ */
+
+#include <linux/module.h>
+#include <linux/bootmem.h>
+#include <linux/memblock.h>
+#include <linux/debugfs.h>
+#include <linux/crashlog.h>
+#include <linux/kmsg_dump.h>
+#include <linux/module.h>
+#include <linux/pfn.h>
+#include <linux/vmalloc.h>
+#include <asm/io.h>
+
+#define CRASHLOG_PAGES 4
+#define CRASHLOG_SIZE (CRASHLOG_PAGES * PAGE_SIZE)
+#define CRASHLOG_MAGIC 0xa1eedead
+
+/*
+ * Start the log at 1M before the end of RAM, as some boot loaders like
+ * to use the end of the RAM for stack usage and other things
+ * If this fails, fall back to using the last part.
+ */
+#define CRASHLOG_OFFSET (1024 * 1024)
+
+struct crashlog_data {
+ u32 magic;
+ u32 len;
+ u8 data[];
+};
+
+static struct debugfs_blob_wrapper crashlog_blob;
+static unsigned long crashlog_addr = 0;
+static struct crashlog_data *crashlog_buf;
+static struct kmsg_dumper dump;
+static bool first = true;
+
+extern struct list_head *crashlog_modules;
+
+static bool crashlog_set_addr(phys_addr_t addr, phys_addr_t size)
+{
+ /* Limit to lower 64 MB to avoid highmem */
+ phys_addr_t limit = 64 * 1024 * 1024;
+
+ if (crashlog_addr)
+ return false;
+
+ if (addr > limit)
+ return false;
+
+ if (addr + size > limit)
+ size = limit - addr;
+
+ crashlog_addr = addr;
+
+ if (addr + size > CRASHLOG_OFFSET)
+ crashlog_addr += size - CRASHLOG_OFFSET;
+
+ return true;
+}
+
+#ifndef CONFIG_NO_BOOTMEM
+void __init crashlog_init_bootmem(bootmem_data_t *bdata)
+{
+ phys_addr_t start, end;
+
+ start = PFN_PHYS(bdata->node_low_pfn);
+ end = PFN_PHYS(bdata->node_min_pfn);
+ if (!crashlog_set_addr(start, end - start))
+ return;
+
+ if (reserve_bootmem(crashlog_addr, CRASHLOG_SIZE, BOOTMEM_EXCLUSIVE) < 0) {
+ printk("Crashlog failed to allocate RAM at address 0x%lx\n",
+ crashlog_addr);
+ crashlog_addr = 0;
+ }
+}
+#endif
+
+#ifdef CONFIG_HAVE_MEMBLOCK
+void __init_memblock crashlog_init_memblock(phys_addr_t addr, phys_addr_t size)
+{
+ if (!crashlog_set_addr(addr, size))
+ return;
+
+ if (memblock_reserve(crashlog_addr, CRASHLOG_SIZE)) {
+ printk("Crashlog failed to allocate RAM at address 0x%lx\n",
+ crashlog_addr);
+ crashlog_addr = 0;
+ }
+}
+#endif
+
+static void __init crashlog_copy(void)
+{
+ if (crashlog_buf->magic != CRASHLOG_MAGIC)
+ return;
+
+ if (!crashlog_buf->len || crashlog_buf->len >
+ CRASHLOG_SIZE - sizeof(*crashlog_buf))
+ return;
+
+ crashlog_blob.size = crashlog_buf->len;
+ crashlog_blob.data = kmemdup(crashlog_buf->data,
+ crashlog_buf->len, GFP_KERNEL);
+
+ debugfs_create_blob("crashlog", 0700, NULL, &crashlog_blob);
+}
+
+static int get_maxlen(void)
+{
+ return CRASHLOG_SIZE - sizeof(*crashlog_buf) - crashlog_buf->len;
+}
+
+static void crashlog_printf(const char *fmt, ...)
+{
+ va_list args;
+ int len = get_maxlen();
+
+ if (!len)
+ return;
+
+ va_start(args, fmt);
+ crashlog_buf->len += vscnprintf(
+ &crashlog_buf->data[crashlog_buf->len],
+ len, fmt, args);
+ va_end(args);
+}
+
+static void crashlog_do_dump(struct kmsg_dumper *dumper,
+ enum kmsg_dump_reason reason)
+{
+ struct timeval tv;
+ struct module *m;
+ char *buf;
+ size_t len;
+
+ if (!first)
+ crashlog_printf("\n===================================\n");
+
+ do_gettimeofday(&tv);
+ crashlog_printf("Time: %lu.%lu\n",
+ (long)tv.tv_sec, (long)tv.tv_usec);
+
+ if (first) {
+ crashlog_printf("Modules:");
+ list_for_each_entry(m, crashlog_modules, list) {
+ crashlog_printf("\t%s@%p+%x", m->name,
+ m->core_layout.base, m->core_layout.size,
+ m->init_layout.base, m->init_layout.size);
+ }
+ crashlog_printf("\n");
+ first = false;
+ }
+
+ buf = (char *)&crashlog_buf->data[crashlog_buf->len];
+
+ kmsg_dump_get_buffer(dumper, true, buf, get_maxlen(), &len);
+
+ crashlog_buf->len += len;
+}
+
+
+int __init crashlog_init_fs(void)
+{
+ struct page *pages[CRASHLOG_PAGES];
+ pgprot_t prot;
+ int i;
+
+ if (!crashlog_addr) {
+ printk("No memory allocated for crashlog\n");
+ return -ENOMEM;
+ }
+
+ printk("Crashlog allocated RAM at address 0x%lx\n", (unsigned long) crashlog_addr);
+ for (i = 0; i < CRASHLOG_PAGES; i++)
+ pages[i] = pfn_to_page((crashlog_addr >> PAGE_SHIFT) + i);
+
+ prot = pgprot_writecombine(PAGE_KERNEL);
+ crashlog_buf = vmap(pages, CRASHLOG_PAGES, VM_MAP, prot);
+
+ crashlog_copy();
+
+ crashlog_buf->magic = CRASHLOG_MAGIC;
+ crashlog_buf->len = 0;
+
+ dump.max_reason = KMSG_DUMP_OOPS;
+ dump.dump = crashlog_do_dump;
+ kmsg_dump_register(&dump);
+
+ return 0;
+}
+module_init(crashlog_init_fs);
--- a/kernel/module.c
+++ b/kernel/module.c
@@ -256,6 +256,9 @@ static void mod_update_bounds(struct mod
#ifdef CONFIG_KGDB_KDB
struct list_head *kdb_modules = &modules; /* kdb needs the list of modules */
#endif /* CONFIG_KGDB_KDB */
+#ifdef CONFIG_CRASHLOG
+struct list_head *crashlog_modules = &modules;
+#endif
static void module_assert_mutex(void)
{
--- a/mm/bootmem.c
+++ b/mm/bootmem.c
@@ -15,6 +15,7 @@
#include <linux/export.h>
#include <linux/kmemleak.h>
#include <linux/range.h>
+#include <linux/crashlog.h>
#include <linux/bug.h>
#include <linux/io.h>
#include <linux/bootmem.h>
@@ -175,6 +176,7 @@ static unsigned long __init free_all_boo
if (!bdata->node_bootmem_map)
return 0;
+ crashlog_init_bootmem(bdata);
map = bdata->node_bootmem_map;
start = bdata->node_min_pfn;
end = bdata->node_low_pfn;
--- a/mm/memblock.c
+++ b/mm/memblock.c
@@ -19,6 +19,7 @@
#include <linux/debugfs.h>
#include <linux/seq_file.h>
#include <linux/memblock.h>
+#include <linux/crashlog.h>
#include <asm/sections.h>
#include <linux/io.h>
@@ -483,6 +484,8 @@ static void __init_memblock memblock_ins
memblock_set_region_node(rgn, nid);
type->cnt++;
type->total_size += size;
+ if (type == &memblock.memory)
+ crashlog_init_memblock(base, size);
}
/**
@@ -522,6 +525,8 @@ int __init_memblock memblock_add_range(s
type->regions[0].flags = flags;
memblock_set_region_node(&type->regions[0], nid);
type->total_size = size;
+ if (type == &memblock.memory)
+ crashlog_init_memblock(base, size);
return 0;
}
repeat:

30
target/linux/generic/pending-4.14/100-MIPS-fix-cache-flushing-for-highmem-pages.patch Normal file
View file

@ -0,0 +1,30 @@
From: Felix Fietkau <nbd@nbd.name>
Subject: MIPS: fix cache flushing for highmem pages
Most cache flush ops were no-op for highmem pages. This led to nasty
segfaults and (in the case of page_address(page) == NULL) kernel
crashes.
Fix this by always flushing highmem pages using kmap/kunmap_atomic
around the actual cache flush. This might be a bit inefficient, but at
least it's stable.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
--- a/arch/mips/mm/cache.c
+++ b/arch/mips/mm/cache.c
@@ -116,6 +116,13 @@ void __flush_anon_page(struct page *page
{
unsigned long addr = (unsigned long) page_address(page);
+ if (PageHighMem(page)) {
+ addr = (unsigned long)kmap_atomic(page);
+ flush_data_cache_page(addr);
+ __kunmap_atomic((void *)addr);
+ return;
+ }
+
if (pages_do_alias(addr, vmaddr)) {
if (page_mapcount(page) && !Page_dcache_dirty(page)) {
void *kaddr;

79
target/linux/generic/pending-4.14/110-ehci_hcd_ignore_oc.patch Normal file
View file

@ -0,0 +1,79 @@
From: Florian Fainelli <florian@openwrt.org>
Subject: USB: EHCI: add ignore_oc flag to disable overcurrent checking
This patch adds an ignore_oc flag which can be set by EHCI controller
not supporting or wanting to disable overcurrent checking. The EHCI
platform data in include/linux/usb/ehci_pdriver.h is also augmented to
take advantage of this new flag.
Signed-off-by: Florian Fainelli <florian@openwrt.org>
---
drivers/usb/host/ehci-hcd.c | 2 +-
drivers/usb/host/ehci-hub.c | 4 ++--
drivers/usb/host/ehci-platform.c | 1 +
drivers/usb/host/ehci.h | 1 +
include/linux/usb/ehci_pdriver.h | 1 +
5 files changed, 6 insertions(+), 3 deletions(-)
--- a/drivers/usb/host/ehci-hcd.c
+++ b/drivers/usb/host/ehci-hcd.c
@@ -651,7 +651,7 @@ static int ehci_run (struct usb_hcd *hcd
"USB %x.%x started, EHCI %x.%02x%s\n",
((ehci->sbrn & 0xf0)>>4), (ehci->sbrn & 0x0f),
temp >> 8, temp & 0xff,
- ignore_oc ? ", overcurrent ignored" : "");
+ (ignore_oc || ehci->ignore_oc) ? ", overcurrent ignored" : "");
ehci_writel(ehci, INTR_MASK,
&ehci->regs->intr_enable); /* Turn On Interrupts */
--- a/drivers/usb/host/ehci-hub.c
+++ b/drivers/usb/host/ehci-hub.c
@@ -646,7 +646,7 @@ ehci_hub_status_data (struct usb_hcd *hc
* always set, seem to clear PORT_OCC and PORT_CSC when writing to
* PORT_POWER; that's surprising, but maybe within-spec.
*/
- if (!ignore_oc)
+ if (!ignore_oc && !ehci->ignore_oc)
mask = PORT_CSC | PORT_PEC | PORT_OCC;
else
mask = PORT_CSC | PORT_PEC;
@@ -1016,7 +1016,7 @@ int ehci_hub_control(
if (temp & PORT_PEC)
status |= USB_PORT_STAT_C_ENABLE << 16;
- if ((temp & PORT_OCC) && !ignore_oc){
+ if ((temp & PORT_OCC) && (!ignore_oc && !ehci->ignore_oc)){
status |= USB_PORT_STAT_C_OVERCURRENT << 16;
/*
--- a/drivers/usb/host/ehci-platform.c
+++ b/drivers/usb/host/ehci-platform.c
@@ -263,6 +263,8 @@ static int ehci_platform_probe(struct pl
hcd->has_tt = 1;
if (pdata->reset_on_resume)
priv->reset_on_resume = true;
+ if (pdata->ignore_oc)
+ ehci->ignore_oc = 1;
#ifndef CONFIG_USB_EHCI_BIG_ENDIAN_MMIO
if (ehci->big_endian_mmio) {
--- a/drivers/usb/host/ehci.h
+++ b/drivers/usb/host/ehci.h
@@ -231,6 +231,7 @@ struct ehci_hcd { /* one per controlle
unsigned frame_index_bug:1; /* MosChip (AKA NetMos) */
unsigned need_oc_pp_cycle:1; /* MPC834X port power */
unsigned imx28_write_fix:1; /* For Freescale i.MX28 */
+ unsigned ignore_oc:1;
/* required for usb32 quirk */
#define OHCI_CTRL_HCFS (3 << 6)
--- a/include/linux/usb/ehci_pdriver.h
+++ b/include/linux/usb/ehci_pdriver.h
@@ -49,6 +49,7 @@ struct usb_ehci_pdata {
unsigned no_io_watchdog:1;
unsigned reset_on_resume:1;
unsigned dma_mask_64:1;
+ unsigned ignore_oc:1;
/* Turn on all power and clocks */
int (*power_on)(struct platform_device *pdev);

82
target/linux/generic/pending-4.14/120-Fix-alloc_node_mem_map-with-ARCH_PFN_OFFSET-calcu.patch Normal file
View file

@ -0,0 +1,82 @@
From: Tobias Wolf <dev-NTEO@vplace.de>
Subject: mm: Fix alloc_node_mem_map with ARCH_PFN_OFFSET calculation
An rt288x (ralink) based router (Belkin F5D8235 v1) does not boot with any
kernel beyond version 4.3 resulting in:
BUG: Bad page state in process swapper pfn:086ac
bisect resulted in:
a1c34a3bf00af2cede839879502e12dc68491ad5 is the first bad commit
commit a1c34a3bf00af2cede839879502e12dc68491ad5
Author: Laura Abbott <laura@labbott.name>
Date: Thu Nov 5 18:48:46 2015 -0800
mm: Don't offset memmap for flatmem
Srinivas Kandagatla reported bad page messages when trying to remove the
bottom 2MB on an ARM based IFC6410 board
BUG: Bad page state in process swapper pfn:fffa8
page:ef7fb500 count:0 mapcount:0 mapping: (null) index:0x0
flags: 0x96640253(locked|error|dirty|active|arch_1|reclaim|mlocked)
page dumped because: PAGE_FLAGS_CHECK_AT_FREE flag(s) set
bad because of flags:
flags: 0x200041(locked|active|mlocked)
Modules linked in:
CPU: 0 PID: 0 Comm: swapper Not tainted 3.19.0-rc3-00007-g412f9ba-dirty
#816
Hardware name: Qualcomm (Flattened Device Tree)
unwind_backtrace
show_stack
dump_stack
bad_page
free_pages_prepare
free_hot_cold_page
__free_pages
free_highmem_page
mem_init
start_kernel
Disabling lock debugging due to kernel taint
[...]
:040000 040000 2de013c372345fd471cd58f0553c9b38b0ef1cc4
0a8156f848733dfa21e16c196dfb6c0a76290709 M mm
This fix for ARM does not account ARCH_PFN_OFFSET for mem_map as later used by
page_to_pfn anymore.
The following output was generated with two hacked in printk statements:
printk("before %p vs. %p or %p\n", mem_map, mem_map - offset, mem_map -
(pgdat->node_start_pfn - ARCH_PFN_OFFSET));
if (page_to_pfn(mem_map) != pgdat->node_start_pfn)
mem_map -= offset + (pgdat->node_start_pfn - ARCH_PFN_OFFSET);
printk("after %p\n", mem_map);
Output:
[ 0.000000] before 8861b280 vs. 8861b280 or 8851b280
[ 0.000000] after 8851b280
As seen in the first line mem_map with subtraction of offset does not equal the
mem_map after subtraction of ARCH_PFN_OFFSET.
After adding the offset of ARCH_PFN_OFFSET as well to mem_map as the
previously calculated offset is zero for the named platform it is able to boot
4.4 and 4.9-rc7 again.
Signed-off-by: Tobias Wolf <dev-NTEO@vplace.de>
---
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -6170,7 +6170,7 @@ static void __ref alloc_node_mem_map(str
mem_map = NODE_DATA(0)->node_mem_map;
#if defined(CONFIG_HAVE_MEMBLOCK_NODE_MAP) || defined(CONFIG_FLATMEM)
if (page_to_pfn(mem_map) != pgdat->node_start_pfn)
- mem_map -= offset;
+ mem_map -= offset + (pgdat->node_start_pfn - ARCH_PFN_OFFSET);
#endif /* CONFIG_HAVE_MEMBLOCK_NODE_MAP */
}
#endif

18
target/linux/generic/pending-4.14/130-add-linux-spidev-compatible-si3210.patch Normal file
View file

@ -0,0 +1,18 @@
From: Giuseppe Lippolis <giu.lippolis@gmail.com>
Subject: Add the linux,spidev compatible in spidev Several device in ramips have this binding in the dts
Signed-off-by: Giuseppe Lippolis <giu.lippolis@gmail.com>
---
drivers/spi/spidev.c | 1 +
1 file changed, 1 insertion(+)
--- a/drivers/spi/spidev.c
+++ b/drivers/spi/spidev.c
@@ -669,6 +669,7 @@ static const struct of_device_id spidev_
{ .compatible = "lineartechnology,ltc2488" },
{ .compatible = "ge,achc" },
{ .compatible = "semtech,sx1301" },
+ { .compatible = "siliconlabs,si3210" },
{},
};
MODULE_DEVICE_TABLE(of, spidev_dt_ids);

20
target/linux/generic/pending-4.14/131-spi-use-gpio_set_value_cansleep-for-setting-chipsele.patch Normal file
View file

@ -0,0 +1,20 @@
From: Felix Fietkau <nbd@nbd.name>
Subject: spi: use gpio_set_value_cansleep for setting chipselect GPIO
Sleeping is safe inside spi_transfer_one_message, and some GPIO chips
need to sleep for setting values
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
--- a/drivers/spi/spi.c
+++ b/drivers/spi/spi.c
@@ -729,7 +729,7 @@ static void spi_set_cs(struct spi_device
enable = !enable;
if (gpio_is_valid(spi->cs_gpio)) {
- gpio_set_value(spi->cs_gpio, !enable);
+ gpio_set_value_cansleep(spi->cs_gpio, !enable);
/* Some SPI masters need both GPIO CS & slave_select */
if ((spi->controller->flags & SPI_MASTER_GPIO_SS) &&
spi->controller->set_cs)

62
target/linux/generic/pending-4.14/140-jffs2-use-.rename2-and-add-RENAME_WHITEOUT-support.patch Normal file
View file

@ -0,0 +1,62 @@
From: Felix Fietkau <nbd@nbd.name>
Subject: jffs2: use .rename2 and add RENAME_WHITEOUT support
It is required for renames on overlayfs
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
--- a/fs/jffs2/dir.c
+++ b/fs/jffs2/dir.c
@@ -756,6 +756,24 @@ static int jffs2_mknod (struct inode *di
return ret;
}
+static int jffs2_whiteout (struct inode *old_dir, struct dentry *old_dentry)
+{
+ struct dentry *wh;
+ int err;
+
+ wh = d_alloc(old_dentry->d_parent, &old_dentry->d_name);
+ if (!wh)
+ return -ENOMEM;
+
+ err = jffs2_mknod(old_dir, wh, S_IFCHR | WHITEOUT_MODE,
+ WHITEOUT_DEV);
+ if (err)
+ return err;
+
+ d_rehash(wh);
+ return 0;
+}
+
static int jffs2_rename (struct inode *old_dir_i, struct dentry *old_dentry,
struct inode *new_dir_i, struct dentry *new_dentry,
unsigned int flags)
@@ -766,7 +784,7 @@ static int jffs2_rename (struct inode *o
uint8_t type;
uint32_t now;
- if (flags & ~RENAME_NOREPLACE)
+ if (flags & ~(RENAME_NOREPLACE|RENAME_WHITEOUT))
return -EINVAL;
/* The VFS will check for us and prevent trying to rename a
@@ -832,9 +850,14 @@ static int jffs2_rename (struct inode *o
if (d_is_dir(old_dentry) && !victim_f)
inc_nlink(new_dir_i);
- /* Unlink the original */
- ret = jffs2_do_unlink(c, JFFS2_INODE_INFO(old_dir_i),
- old_dentry->d_name.name, old_dentry->d_name.len, NULL, now);
+ if (flags & RENAME_WHITEOUT)
+ /* Replace with whiteout */
+ ret = jffs2_whiteout(old_dir_i, old_dentry);
+ else
+ /* Unlink the original */
+ ret = jffs2_do_unlink(c, JFFS2_INODE_INFO(old_dir_i),
+ old_dentry->d_name.name,
+ old_dentry->d_name.len, NULL, now);
/* We don't touch inode->i_nlink */

73
target/linux/generic/pending-4.14/141-jffs2-add-RENAME_EXCHANGE-support.patch Normal file
View file

@ -0,0 +1,73 @@
From: Felix Fietkau <nbd@nbd.name>
Subject: jffs2: add RENAME_EXCHANGE support
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
--- a/fs/jffs2/dir.c
+++ b/fs/jffs2/dir.c
@@ -781,18 +781,31 @@ static int jffs2_rename (struct inode *o
int ret;
struct jffs2_sb_info *c = JFFS2_SB_INFO(old_dir_i->i_sb);
struct jffs2_inode_info *victim_f = NULL;
+ struct inode *fst_inode = d_inode(old_dentry);
+ struct inode *snd_inode = d_inode(new_dentry);
uint8_t type;
uint32_t now;
- if (flags & ~(RENAME_NOREPLACE|RENAME_WHITEOUT))
+ if (flags & ~(RENAME_NOREPLACE|RENAME_WHITEOUT|RENAME_EXCHANGE))
return -EINVAL;
+ if ((flags & RENAME_EXCHANGE) && (old_dir_i != new_dir_i)) {
+ if (S_ISDIR(fst_inode->i_mode) && !S_ISDIR(snd_inode->i_mode)) {
+ inc_nlink(new_dir_i);
+ drop_nlink(old_dir_i);
+ }
+ else if (!S_ISDIR(fst_inode->i_mode) && S_ISDIR(snd_inode->i_mode)) {
+ drop_nlink(new_dir_i);
+ inc_nlink(old_dir_i);
+ }
+ }
+
/* The VFS will check for us and prevent trying to rename a
* file over a directory and vice versa, but if it's a directory,
* the VFS can't check whether the victim is empty. The filesystem
* needs to do that for itself.
*/
- if (d_really_is_positive(new_dentry)) {
+ if (d_really_is_positive(new_dentry) && !(flags & RENAME_EXCHANGE)) {
victim_f = JFFS2_INODE_INFO(d_inode(new_dentry));
if (d_is_dir(new_dentry)) {
struct jffs2_full_dirent *fd;
@@ -827,7 +840,7 @@ static int jffs2_rename (struct inode *o
if (ret)
return ret;
- if (victim_f) {
+ if (victim_f && !(flags & RENAME_EXCHANGE)) {
/* There was a victim. Kill it off nicely */
if (d_is_dir(new_dentry))
clear_nlink(d_inode(new_dentry));
@@ -853,6 +866,12 @@ static int jffs2_rename (struct inode *o
if (flags & RENAME_WHITEOUT)
/* Replace with whiteout */
ret = jffs2_whiteout(old_dir_i, old_dentry);
+ else if (flags & RENAME_EXCHANGE)
+ /* Replace the original */
+ ret = jffs2_do_link(c, JFFS2_INODE_INFO(old_dir_i),
+ d_inode(new_dentry)->i_ino, type,
+ old_dentry->d_name.name, old_dentry->d_name.len,
+ now);
else
/* Unlink the original */
ret = jffs2_do_unlink(c, JFFS2_INODE_INFO(old_dir_i),
@@ -884,7 +903,7 @@ static int jffs2_rename (struct inode *o
return ret;
}
- if (d_is_dir(old_dentry))
+ if (d_is_dir(old_dentry) && !(flags & RENAME_EXCHANGE))
drop_nlink(old_dir_i);
new_dir_i->i_mtime = new_dir_i->i_ctime = old_dir_i->i_mtime = old_dir_i->i_ctime = ITIME(now);

43
target/linux/generic/pending-4.14/150-bridge_allow_receiption_on_disabled_port.patch Normal file
View file

@ -0,0 +1,43 @@
From: Stephen Hemminger <stephen@networkplumber.org>
Subject: bridge: allow receiption on disabled port
When an ethernet device is enslaved to a bridge, and the bridge STP
detects loss of carrier (or operational state down), then normally
packet receiption is blocked.
This breaks control applications like WPA which maybe expecting to
receive packets to negotiate to bring link up. The bridge needs to
block forwarding packets from these disabled ports, but there is no
hard requirement to not allow local packet delivery.
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
--- a/net/bridge/br_input.c
+++ b/net/bridge/br_input.c
@@ -238,7 +238,8 @@ static int br_handle_local_finish(struct
{
struct net_bridge_port *p = br_port_get_rcu(skb->dev);
- __br_handle_local_finish(skb);
+ if (p->state != BR_STATE_DISABLED)
+ __br_handle_local_finish(skb);
BR_INPUT_SKB_CB(skb)->brdev = p->br->dev;
br_pass_frame_up(skb);
@@ -326,6 +327,15 @@ rx_handler_result_t br_handle_frame(stru
forward:
switch (p->state) {
+ case BR_STATE_DISABLED:
+ if (ether_addr_equal(p->br->dev->dev_addr, dest))
+ skb->pkt_type = PACKET_HOST;
+
+ NF_HOOK(NFPROTO_BRIDGE, NF_BR_PRE_ROUTING,
+ dev_net(skb->dev), NULL, skb, skb->dev, NULL,
+ br_handle_local_finish);
+ break;
+
case BR_STATE_FORWARDING:
rhook = rcu_dereference(br_should_route_hook);
if (rhook) {

161
target/linux/generic/pending-4.14/160-mtd-part-add-generic-parsing-of-linux-part-probe.patch Normal file
View file

@ -0,0 +1,161 @@
From: Hauke Mehrtens <hauke@hauke-m.de>
Subject: mtd: part: add generic parsing of linux,part-probe
This moves the linux,part-probe device tree parsing code from
physmap_of.c to mtdpart.c. Now all drivers can use this feature by just
providing a reference to their device tree node in struct
mtd_part_parser_data.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
---
Documentation/devicetree/bindings/mtd/nand.txt | 16 +++++++++
drivers/mtd/maps/physmap_of.c | 46 +-------------------------
drivers/mtd/mtdpart.c | 45 +++++++++++++++++++++++++
3 files changed, 62 insertions(+), 45 deletions(-)
--- a/Documentation/devicetree/bindings/mtd/nand.txt
+++ b/Documentation/devicetree/bindings/mtd/nand.txt
@@ -44,6 +44,22 @@ Optional NAND chip properties:
used by the upper layers, and you want to make your NAND
as reliable as possible.
+- linux,part-probe: list of name as strings of the partition parser
+ which should be used to parse the partition table.
+ They will be tried in the specified ordering and
+ the next one will be used if the previous one
+ failed.
+
+ Example: linux,part-probe = "cmdlinepart", "ofpart";
+
+ This is also the default value, which will be used
+ if this attribute is not specified. It could be
+ that the flash driver in use overwrote the default
+ value and uses some other default.
+
+ Possible values are: bcm47xxpart, afs, ar7part,
+ ofoldpart, ofpart, bcm63xxpart, RedBoot, cmdlinepart
+
The ECC strength and ECC step size properties define the correction capability
of a controller. Together, they say a controller can correct "{strength} bit
errors per {size} bytes".
--- a/drivers/mtd/maps/physmap_of_core.c
+++ b/drivers/mtd/maps/physmap_of_core.c
@@ -114,37 +114,9 @@ static struct mtd_info *obsolete_probe(s
static const char * const part_probe_types_def[] = {
"cmdlinepart", "RedBoot", "ofpart", "ofoldpart", NULL };
-static const char * const *of_get_probes(struct device_node *dp)
-{
- const char **res;
- int count;
-
- count = of_property_count_strings(dp, "linux,part-probe");
- if (count < 0)
- return part_probe_types_def;
-
- res = kzalloc((count + 1) * sizeof(*res), GFP_KERNEL);
- if (!res)
- return NULL;
-
- count = of_property_read_string_array(dp, "linux,part-probe", res,
- count);
- if (count < 0)
- return NULL;
-
- return res;
-}
-
-static void of_free_probes(const char * const *probes)
-{
- if (probes != part_probe_types_def)
- kfree(probes);
-}
-
static const struct of_device_id of_flash_match[];
static int of_flash_probe(struct platform_device *dev)
{
- const char * const *part_probe_types;
const struct of_device_id *match;
struct device_node *dp = dev->dev.of_node;
struct resource res;
@@ -310,14 +282,8 @@ static int of_flash_probe(struct platfor
info->cmtd->dev.parent = &dev->dev;
mtd_set_of_node(info->cmtd, dp);
- part_probe_types = of_get_probes(dp);
- if (!part_probe_types) {
- err = -ENOMEM;
- goto err_out;
- }
- mtd_device_parse_register(info->cmtd, part_probe_types, NULL,
+ mtd_device_parse_register(info->cmtd, part_probe_types_def, NULL,
NULL, 0);
- of_free_probes(part_probe_types);
kfree(mtd_list);
--- a/drivers/mtd/mtdpart.c
+++ b/drivers/mtd/mtdpart.c
@@ -29,6 +29,7 @@
#include <linux/kmod.h>
#include <linux/mtd/mtd.h>
#include <linux/mtd/partitions.h>
+#include <linux/of.h>
#include <linux/err.h>
#include "mtdcore.h"
@@ -863,6 +864,32 @@ void deregister_mtd_parser(struct mtd_pa
EXPORT_SYMBOL_GPL(deregister_mtd_parser);
/*
+ * Parses the linux,part-probe device tree property.
+ * When a non null value is returned it has to be freed with kfree() by
+ * the caller.
+ */
+static const char * const *of_get_probes(struct device_node *dp)
+{
+ const char **res;
+ int count;
+
+ count = of_property_count_strings(dp, "linux,part-probe");
+ if (count < 0)
+ return NULL;
+
+ res = kzalloc((count + 1) * sizeof(*res), GFP_KERNEL);
+ if (!res)
+ return NULL;
+
+ count = of_property_read_string_array(dp, "linux,part-probe", res,
+ count);
+ if (count < 0)
+ return NULL;
+
+ return res;
+}
+
+/*
* Do not forget to update 'parse_mtd_partitions()' kerneldoc comment if you
* are changing this array!
*/
@@ -920,6 +947,13 @@ int parse_mtd_partitions(struct mtd_info
{
struct mtd_part_parser *parser;
int ret, err = 0;
+ const char *const *types_of = NULL;
+
+ if (mtd_get_of_node(master)) {
+ types_of = of_get_probes(mtd_get_of_node(master));
+ if (types_of != NULL)
+ types = types_of;
+ }
if (!types)
types = default_mtd_part_types;
@@ -945,6 +979,7 @@ int parse_mtd_partitions(struct mtd_info
if (ret < 0 && !err)
err = ret;
}
+ kfree(types_of);
return err;
}

70
target/linux/generic/pending-4.14/180-net-phy-at803x-add-support-for-AT8032.patch Normal file
View file

@ -0,0 +1,70 @@
From: Felix Fietkau <nbd@nbd.name>
Subject: net: phy: at803x: add support for AT8032
Like AT8030, this PHY needs the GPIO reset workaround
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
--- a/drivers/net/phy/at803x.c
+++ b/drivers/net/phy/at803x.c
@@ -62,6 +62,7 @@
#define ATH8030_PHY_ID 0x004dd076
#define ATH8031_PHY_ID 0x004dd074
+#define ATH8032_PHY_ID 0x004dd023
#define ATH8035_PHY_ID 0x004dd072
#define AT803X_PHY_ID_MASK 0xffffffef
@@ -260,7 +261,8 @@ static int at803x_probe(struct phy_devic
if (!priv)
return -ENOMEM;
- if (phydev->drv->phy_id != ATH8030_PHY_ID)
+ if (phydev->drv->phy_id != ATH8030_PHY_ID &&
+ phydev->drv->phy_id != ATH8032_PHY_ID)
goto does_not_require_reset_workaround;
gpiod_reset = devm_gpiod_get_optional(dev, "reset", GPIOD_OUT_LOW);
@@ -336,7 +338,7 @@ static void at803x_link_change_notify(st
struct at803x_priv *priv = phydev->priv;
/*
- * Conduct a hardware reset for AT8030 every time a link loss is
+ * Conduct a hardware reset for AT8030/2 every time a link loss is
* signalled. This is necessary to circumvent a hardware bug that
* occurs when the cable is unplugged while TX packets are pending
* in the FIFO. In such cases, the FIFO enters an error mode it
@@ -448,6 +450,24 @@ static struct phy_driver at803x_driver[]
.aneg_done = at803x_aneg_done,
.ack_interrupt = &at803x_ack_interrupt,
.config_intr = &at803x_config_intr,
+}, {
+ /* ATHEROS 8032 */
+ .phy_id = ATH8032_PHY_ID,
+ .name = "Atheros 8032 ethernet",
+ .phy_id_mask = 0xffffffef,
+ .probe = at803x_probe,
+ .config_init = at803x_config_init,
+ .link_change_notify = at803x_link_change_notify,
+ .set_wol = at803x_set_wol,
+ .get_wol = at803x_get_wol,
+ .suspend = at803x_suspend,
+ .resume = at803x_resume,
+ .features = PHY_BASIC_FEATURES,
+ .flags = PHY_HAS_INTERRUPT,
+ .config_aneg = genphy_config_aneg,
+ .read_status = genphy_read_status,
+ .ack_interrupt = at803x_ack_interrupt,
+ .config_intr = at803x_config_intr,
} };
module_phy_driver(at803x_driver);
@@ -455,6 +475,7 @@ module_phy_driver(at803x_driver);
static struct mdio_device_id __maybe_unused atheros_tbl[] = {
{ ATH8030_PHY_ID, AT803X_PHY_ID_MASK },
{ ATH8031_PHY_ID, AT803X_PHY_ID_MASK },
+ { ATH8032_PHY_ID, AT803X_PHY_ID_MASK },
{ ATH8035_PHY_ID, AT803X_PHY_ID_MASK },
{ }
};

43
target/linux/generic/pending-4.14/190-2-5-e1000e-Fix-wrong-comment-related-to-link-detection.patch Normal file
View file

@ -0,0 +1,43 @@
From patchwork Fri Jul 21 18:36:24 2017
Content-Type: text/plain; charset="utf-8"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Subject: [2/5] e1000e: Fix wrong comment related to link detection
From: Benjamin Poirier <bpoirier@suse.com>
X-Patchwork-Id: 9857489
Message-Id: <20170721183627.13373-2-bpoirier@suse.com>
To: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Cc: Lennart Sorensen <lsorense@csclub.uwaterloo.ca>,
intel-wired-lan@lists.osuosl.org, netdev@vger.kernel.org,
linux-kernel@vger.kernel.org
Date: Fri, 21 Jul 2017 11:36:24 -0700
Reading e1000e_check_for_copper_link() shows that get_link_status is set to
false after link has been detected. Therefore, it stays TRUE until then.
Signed-off-by: Benjamin Poirier <bpoirier@suse.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
---
drivers/net/ethernet/intel/e1000e/netdev.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
--- a/drivers/net/ethernet/intel/e1000e/netdev.c
+++ b/drivers/net/ethernet/intel/e1000e/netdev.c
@@ -5096,7 +5096,7 @@ static bool e1000e_has_link(struct e1000
/* get_link_status is set on LSC (link status) interrupt or
* Rx sequence error interrupt. get_link_status will stay
- * false until the check_for_link establishes link
+ * true until the check_for_link establishes link
* for copper adapters ONLY
*/
switch (hw->phy.media_type) {
@@ -5114,7 +5114,7 @@ static bool e1000e_has_link(struct e1000
break;
case e1000_media_type_internal_serdes:
ret_val = hw->mac.ops.check_for_link(hw);
- link_active = adapter->hw.mac.serdes_has_link;
+ link_active = hw->mac.serdes_has_link;
break;
default:
case e1000_media_type_unknown:

32
target/linux/generic/pending-4.14/201-extra_optimization.patch Normal file
View file

@ -0,0 +1,32 @@
From: Felix Fietkau <nbd@nbd.name>
Subject: Upgrade to Linux 2.6.19
- Includes large parts of the patch from #1021 by dpalffy
- Includes RB532 NAND driver changes by n0-1
[john@phrozen.org: feix will add this to his upstream queue]
lede-commit: bff468813f78f81e36ebb2a3f4354de7365e640f
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
Makefile | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
--- a/Makefile
+++ b/Makefile
@@ -637,12 +637,12 @@ KBUILD_CFLAGS += $(call cc-disable-warni
ifdef CONFIG_CC_OPTIMIZE_FOR_SIZE
KBUILD_CFLAGS += $(call cc-option,-Oz,-Os)
-KBUILD_CFLAGS += $(call cc-disable-warning,maybe-uninitialized,)
+KBUILD_CFLAGS += $(call cc-disable-warning,maybe-uninitialized,) $(EXTRA_OPTIMIZATION)
else
ifdef CONFIG_PROFILE_ALL_BRANCHES
-KBUILD_CFLAGS += -O2 $(call cc-disable-warning,maybe-uninitialized,)
+KBUILD_CFLAGS += -O2 $(call cc-disable-warning,maybe-uninitialized,) $(EXTRA_OPTIMIZATION)
else
-KBUILD_CFLAGS += -O2
+KBUILD_CFLAGS += -O2 -fno-reorder-blocks -fno-tree-ch $(EXTRA_OPTIMIZATION)
endif
endif

119
target/linux/generic/pending-4.14/203-kallsyms_uncompressed.patch Normal file
View file

@ -0,0 +1,119 @@
From: Felix Fietkau <nbd@nbd.name>
Subject: kernel: add a config option for keeping the kallsyms table uncompressed, saving ~9kb kernel size after lzma on ar71xx
[john@phrozen.org: added to my upstream queue 30.12.2016]
lede-commit: e0e3509b5ce2ccf93d4d67ea907613f5f7ec2eed
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
init/Kconfig | 11 +++++++++++
kernel/kallsyms.c | 8 ++++++++
scripts/kallsyms.c | 12 ++++++++++++
scripts/link-vmlinux.sh | 4 ++++
4 files changed, 35 insertions(+)
--- a/init/Kconfig
+++ b/init/Kconfig
@@ -1081,6 +1081,17 @@ config SYSCTL_ARCH_UNALIGN_ALLOW
the unaligned access emulation.
see arch/parisc/kernel/unaligned.c for reference
+config KALLSYMS_UNCOMPRESSED
+ bool "Keep kallsyms uncompressed"
+ depends on KALLSYMS
+ help
+ Normally kallsyms contains compressed symbols (using a token table),
+ reducing the uncompressed kernel image size. Keeping the symbol table
+ uncompressed significantly improves the size of this part in compressed
+ kernel images.
+
+ Say N unless you need compressed kernel images to be small.
+
config HAVE_PCSPKR_PLATFORM
bool
--- a/kernel/kallsyms.c
+++ b/kernel/kallsyms.c
@@ -108,6 +108,11 @@ static unsigned int kallsyms_expand_symb
* For every byte on the compressed symbol data, copy the table
* entry for that byte.
*/
+#ifdef CONFIG_KALLSYMS_UNCOMPRESSED
+ memcpy(result, data + 1, len - 1);
+ result += len - 1;
+ len = 0;
+#endif
while (len) {
tptr = &kallsyms_token_table[kallsyms_token_index[*data]];
data++;
@@ -140,6 +145,9 @@ tail:
*/
static char kallsyms_get_symbol_type(unsigned int off)
{
+#ifdef CONFIG_KALLSYMS_UNCOMPRESSED
+ return kallsyms_names[off + 1];
+#endif
/*
* Get just the first code, look it up in the token table,
* and return the first char from this token.
--- a/scripts/kallsyms.c
+++ b/scripts/kallsyms.c
@@ -61,6 +61,7 @@ static struct addr_range percpu_range =
static struct sym_entry *table;
static unsigned int table_size, table_cnt;
static int all_symbols = 0;
+static int uncompressed = 0;
static int absolute_percpu = 0;
static char symbol_prefix_char = '\0';
static int base_relative = 0;
@@ -457,6 +458,9 @@ static void write_src(void)
free(markers);
+ if (uncompressed)
+ return;
+
output_label("kallsyms_token_table");
off = 0;
for (i = 0; i < 256; i++) {
@@ -515,6 +519,9 @@ static void *find_token(unsigned char *s
{
int i;
+ if (uncompressed)
+ return NULL;
+
for (i = 0; i < len - 1; i++) {
if (str[i] == token[0] && str[i+1] == token[1])
return &str[i];
@@ -587,6 +594,9 @@ static void optimize_result(void)
{
int i, best;
+ if (uncompressed)
+ return;
+
/* using the '\0' symbol last allows compress_symbols to use standard
* fast string functions */
for (i = 255; i >= 0; i--) {
@@ -775,6 +785,8 @@ int main(int argc, char **argv)
symbol_prefix_char = *p;
} else if (strcmp(argv[i], "--base-relative") == 0)
base_relative = 1;
+ else if (strcmp(argv[i], "--uncompressed") == 0)
+ uncompressed = 1;
else
usage();
}
--- a/scripts/link-vmlinux.sh
+++ b/scripts/link-vmlinux.sh
@@ -164,6 +164,10 @@ kallsyms()
kallsymopt="${kallsymopt} --base-relative"
fi
+ if [ -n "${CONFIG_KALLSYMS_UNCOMPRESSED}" ]; then
+ kallsymopt="${kallsymopt} --uncompressed"
+ fi
+
local aflags="${KBUILD_AFLAGS} ${KBUILD_AFLAGS_KERNEL} \
${NOSTDINC_FLAGS} ${LINUXINCLUDE} ${KBUILD_CPPFLAGS}"

45
target/linux/generic/pending-4.14/205-backtrace_module_info.patch Normal file
View file

@ -0,0 +1,45 @@
From: Felix Fietkau <nbd@nbd.name>
Subject: kernel: when KALLSYMS is disabled, print module address + size for matching backtrace entries
[john@phrozen.org: felix will add this to his upstream queue]
lede-commit 53827cdc824556cda910b23ce5030c363b8f1461
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
lib/vsprintf.c | 15 +++++++++++----
1 file changed, 11 insertions(+), 4 deletions(-)
--- a/lib/vsprintf.c
+++ b/lib/vsprintf.c
@@ -670,8 +670,10 @@ char *symbol_string(char *buf, char *end
struct printf_spec spec, const char *fmt)
{
unsigned long value;
-#ifdef CONFIG_KALLSYMS
char sym[KSYM_SYMBOL_LEN];
+#ifndef CONFIG_KALLSYMS
+ struct module *mod;
+ int len;
#endif
if (fmt[1] == 'R')
@@ -685,11 +687,16 @@ char *symbol_string(char *buf, char *end
sprint_symbol(sym, value);
else
sprint_symbol_no_offset(sym, value);
-
- return string(buf, end, sym, spec);
#else
- return special_hex_number(buf, end, value, sizeof(void *));
+ len = snprintf(sym, sizeof(sym), "0x%lx", value);
+
+ mod = __module_address(value);
+ if (mod)
+ snprintf(sym + len, sizeof(sym) - len, " [%s@%p+0x%x]",
+ mod->name, mod->core_layout.base,
+ mod->core_layout.size);
#endif
+ return string(buf, end, sym, spec);
}
static noinline_for_stack

23
target/linux/generic/pending-4.14/206-mips-disable-vdso.patch Normal file
View file

@ -0,0 +1,23 @@
From: Felix Fietkau <nbd@nbd.name>
Subject: kernel: disable MIPS VDSO by default until the cache issues have been resolved
lede-commit: 1185e645a773c86aa88cf04d0e2911dc62eb43f5
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
arch/mips/vdso/Makefile | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
--- a/arch/mips/vdso/Makefile
+++ b/arch/mips/vdso/Makefile
@@ -30,9 +30,9 @@ aflags-vdso := $(ccflags-vdso) \
ifndef CONFIG_CPU_MIPSR6
ifeq ($(call ld-ifversion, -lt, 225000000, y),y)
$(warning MIPS VDSO requires binutils >= 2.25)
- obj-vdso-y := $(filter-out gettimeofday.o, $(obj-vdso-y))
- ccflags-vdso += -DDISABLE_MIPS_VDSO
endif
+ obj-vdso-y := $(filter-out gettimeofday.o, $(obj-vdso-y))
+ ccflags-vdso += -DDISABLE_MIPS_VDSO
endif
# VDSO linker flags.

46
target/linux/generic/pending-4.14/240-remove-unsane-filenames-from-deps_initramfs-list.patch Normal file
View file

@ -0,0 +1,46 @@
From: Gabor Juhos <juhosg@openwrt.org>
Subject: usr: sanitize deps_initramfs list
If any filename in the intramfs dependency
list contains a colon, that causes a kernel
build error like this:
/devel/openwrt/build_dir/linux-ar71xx_generic/linux-3.6.6/usr/Makefile:58: *** multiple target patterns. Stop.
make[5]: *** [usr] Error 2
Fix it by removing such filenames from the
deps_initramfs list.
Signed-off-by: Gabor Juhos <juhosg@openwrt.org>
---
usr/Makefile | 8 +++++---
1 file changed, 5 insertions(+), 3 deletions(-)
--- a/usr/Makefile
+++ b/usr/Makefile
@@ -39,20 +39,22 @@ ifneq ($(wildcard $(obj)/$(datafile_d_y)
include $(obj)/$(datafile_d_y)
endif
+deps_initramfs_sane := $(foreach v,$(deps_initramfs),$(if $(findstring :,$(v)),,$(v)))
+
quiet_cmd_initfs = GEN $@
cmd_initfs = $(initramfs) -o $@ $(ramfs-args) $(ramfs-input)
targets := $(datafile_y)
# do not try to update files included in initramfs
-$(deps_initramfs): ;
+$(deps_initramfs_sane): ;
-$(deps_initramfs): klibcdirs
+$(deps_initramfs_sane): klibcdirs
# We rebuild initramfs_data.cpio if:
# 1) Any included file is newer then initramfs_data.cpio
# 2) There are changes in which files are included (added or deleted)
# 3) If gen_init_cpio are newer than initramfs_data.cpio
# 4) arguments to gen_initramfs.sh changes
-$(obj)/$(datafile_y): $(obj)/gen_init_cpio $(deps_initramfs) klibcdirs
+$(obj)/$(datafile_y): $(obj)/gen_init_cpio $(deps_initramfs_sane) klibcdirs
$(Q)$(initramfs) -l $(ramfs-input) > $(obj)/$(datafile_d_y)
$(call if_changed,initfs)

20
target/linux/generic/pending-4.14/261-enable_wilink_platform_without_drivers.patch Normal file
View file

@ -0,0 +1,20 @@
From: Imre Kaloz <kaloz@openwrt.org>
Subject: [PATCH] hack: net: wireless: make the wl12xx glue code available with
compat-wireless, too
Signed-off-by: Imre Kaloz <kaloz@openwrt.org>
---
drivers/net/wireless/ti/Kconfig | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
--- a/drivers/net/wireless/ti/Kconfig
+++ b/drivers/net/wireless/ti/Kconfig
@@ -19,7 +19,7 @@ source "drivers/net/wireless/ti/wlcore/K
config WILINK_PLATFORM_DATA
bool "TI WiLink platform data"
- depends on WLCORE_SDIO || WL1251_SDIO
+ depends on WLCORE_SDIO || WL1251_SDIO || ARCH_OMAP2PLUS
default y
---help---
Small platform data bit needed to pass data to the sdio modules.

32
target/linux/generic/pending-4.14/270-uapi-kernel.h-glibc-specific-inclusion-of-sysinfo.h.patch Normal file
View file

@ -0,0 +1,32 @@
From: David Heidelberger <david.heidelberger@ixit.cz>
Subject: uapi/kernel.h: glibc specific inclusion of sysinfo.h
including sysinfo.h from kernel.h makes no sense whatsoever,
but removing it breaks glibc's userspace header,
which includes kernel.h instead of sysinfo.h from their sys/sysinfo.h.
this seems to be a historical mistake.
on musl, including any header that uses kernel.h directly or indirectly
plus sys/sysinfo.h will produce a compile error due to redefinition of
struct sysinfo from sys/sysinfo.h.
so for now, only include it on glibc or when including from kernel
in order not to break their headers.
Signed-off-by: John Spencer <maillist-linux@barfooze.de>
Signed-off-by: David Heidelberger <david.heidelberger@ixit.cz>
Signed-off-by: Jonas Gorski <jogo@openwrt.org>
---
include/uapi/linux/kernel.h | 2 ++
1 file changed, 2 insertions(+)
--- a/include/uapi/linux/kernel.h
+++ b/include/uapi/linux/kernel.h
@@ -2,7 +2,9 @@
#ifndef _UAPI_LINUX_KERNEL_H
#define _UAPI_LINUX_KERNEL_H
+#if defined(__KERNEL__) || defined( __GLIBC__)
#include <linux/sysinfo.h>
+#endif
/*
* 'kernel.h' contains some often-used function prototypes etc

105
target/linux/generic/pending-4.14/271-uapi-libc-compat.h-do-not-rely-on-__GLIBC__.patch Normal file
View file

@ -0,0 +1,105 @@
From: David Heidelberger <david.heidelberger@ixit.cz>
Subject: uapi/libc-compat.h: do not rely on __GLIBC__
Musl provides the same structs as glibc, but does not provide a define to
allow its detection. Since the absence of __GLIBC__ also can mean that it
is included from the kernel, change the __GLIBC__ detection to
!__KERNEL__, which should always be true when included from userspace.
Signed-off-by: John Spencer <maillist-linux@barfooze.de>
Tested-by: David Heidelberger <david.heidelberger@ixit.cz>
Signed-off-by: Jonas Gorski <jogo@openwrt.org>
---
include/uapi/linux/libc-compat.h | 18 +++++++++---------
1 file changed, 9 insertions(+), 9 deletions(-)
--- a/include/uapi/linux/libc-compat.h
+++ b/include/uapi/linux/libc-compat.h
@@ -49,13 +49,13 @@
#ifndef _UAPI_LIBC_COMPAT_H
#define _UAPI_LIBC_COMPAT_H
-/* We have included glibc headers... */
-#if defined(__GLIBC__)
+/* We have included libc headers... */
+#if !defined(__KERNEL__)
-/* Coordinate with glibc net/if.h header. */
-#if defined(_NET_IF_H) && defined(__USE_MISC)
+/* Coordinate with libc net/if.h header. */
+#if defined(_NET_IF_H) && (!defined(__GLIBC__) || defined(__USE_MISC))
-/* GLIBC headers included first so don't define anything
+/* LIBC headers included first so don't define anything
* that would already be defined. */
#define __UAPI_DEF_IF_IFCONF 0
@@ -66,7 +66,11 @@
#define __UAPI_DEF_IF_NET_DEVICE_FLAGS 0
/* For the future if glibc adds IFF_LOWER_UP, IFF_DORMANT and IFF_ECHO */
#ifndef __UAPI_DEF_IF_NET_DEVICE_FLAGS_LOWER_UP_DORMANT_ECHO
+#ifdef __GLIBC__
#define __UAPI_DEF_IF_NET_DEVICE_FLAGS_LOWER_UP_DORMANT_ECHO 1
+#else
+#define __UAPI_DEF_IF_NET_DEVICE_FLAGS_LOWER_UP_DORMANT_ECHO 0
+#endif
#endif /* __UAPI_DEF_IF_NET_DEVICE_FLAGS_LOWER_UP_DORMANT_ECHO */
#else /* _NET_IF_H */
@@ -86,10 +90,10 @@
#endif /* _NET_IF_H */
-/* Coordinate with glibc netinet/in.h header. */
+/* Coordinate with libc netinet/in.h header. */
#if defined(_NETINET_IN_H)
-/* GLIBC headers included first so don't define anything
+/* LIBC headers included first so don't define anything
* that would already be defined. */
#define __UAPI_DEF_IN_ADDR 0
#define __UAPI_DEF_IN_IPPROTO 0
@@ -103,7 +107,7 @@
* if the glibc code didn't define them. This guard matches
* the guard in glibc/inet/netinet/in.h which defines the
* additional in6_addr macros e.g. s6_addr16, and s6_addr32. */
-#if defined(__USE_MISC) || defined (__USE_GNU)
+#if !defined(__GLIBC__) || defined(__USE_MISC) || defined (__USE_GNU)
#define __UAPI_DEF_IN6_ADDR_ALT 0
#else
#define __UAPI_DEF_IN6_ADDR_ALT 1
@@ -118,7 +122,7 @@
#else
/* Linux headers included first, and we must define everything
- * we need. The expectation is that glibc will check the
+ * we need. The expectation is that the libc will check the
* __UAPI_DEF_* defines and adjust appropriately. */
#define __UAPI_DEF_IN_ADDR 1
#define __UAPI_DEF_IN_IPPROTO 1
@@ -128,7 +132,7 @@
#define __UAPI_DEF_IN_CLASS 1
#define __UAPI_DEF_IN6_ADDR 1
-/* We unconditionally define the in6_addr macros and glibc must
+/* We unconditionally define the in6_addr macros and the libc must
* coordinate. */
#define __UAPI_DEF_IN6_ADDR_ALT 1
#define __UAPI_DEF_SOCKADDR_IN6 1
@@ -169,7 +173,7 @@
/* If we did not see any headers from any supported C libraries,
* or we are being included in the kernel, then define everything
* that we need. */
-#else /* !defined(__GLIBC__) */
+#else /* defined(__KERNEL__) */
/* Definitions for if.h */
#define __UAPI_DEF_IF_IFCONF 1
@@ -209,6 +213,6 @@
/* Definitions for xattr.h */
#define __UAPI_DEF_XATTR 1
-#endif /* __GLIBC__ */
+#endif /* __KERNEL__ */
#endif /* _UAPI_LIBC_COMPAT_H */

65
target/linux/generic/pending-4.14/272-uapi-if_ether.h-prevent-redefinition-of-struct-ethhd.patch Normal file
View file

@ -0,0 +1,65 @@
From: David Heidelberger <david.heidelberger@ixit.cz>
Subject: uapi/if_ether.h: prevent redefinition of struct ethhdr
Musl provides its own ethhdr struct definition. Add a guard to prevent
its definition of the appropriate musl header has already been included.
Signed-off-by: John Spencer <maillist-linux@barfooze.de>
Tested-by: David Heidelberger <david.heidelberger@ixit.cz>
Signed-off-by: Jonas Gorski <jogo@openwrt.org>
---
include/uapi/linux/if_ether.h | 3 +++
include/uapi/linux/libc-compat.h | 11 +++++++++++
2 files changed, 14 insertions(+)
--- a/include/uapi/linux/if_ether.h
+++ b/include/uapi/linux/if_ether.h
@@ -23,6 +23,7 @@
#define _UAPI_LINUX_IF_ETHER_H
#include <linux/types.h>
+#include <linux/libc-compat.h>
/*
* IEEE 802.3 Ethernet magic constants. The frame sizes omit the preamble
@@ -149,11 +150,13 @@
* This is an Ethernet frame header.
*/
+#if __UAPI_DEF_ETHHDR
struct ethhdr {
unsigned char h_dest[ETH_ALEN]; /* destination eth addr */
unsigned char h_source[ETH_ALEN]; /* source ether addr */
__be16 h_proto; /* packet type ID field */
} __attribute__((packed));
+#endif
#endif /* _UAPI_LINUX_IF_ETHER_H */
--- a/include/uapi/linux/libc-compat.h
+++ b/include/uapi/linux/libc-compat.h
@@ -90,6 +90,14 @@
#endif /* _NET_IF_H */
+/* musl defines the ethhdr struct itself in its netinet/if_ether.h.
+ * Glibc just includes the kernel header and uses a different guard. */
+#if defined(_NETINET_IF_ETHER_H)
+#define __UAPI_DEF_ETHHDR 0
+#else
+#define __UAPI_DEF_ETHHDR 1
+#endif
+
/* Coordinate with libc netinet/in.h header. */
#if defined(_NETINET_IN_H)
@@ -185,6 +193,9 @@
/* For the future if glibc adds IFF_LOWER_UP, IFF_DORMANT and IFF_ECHO */
#define __UAPI_DEF_IF_NET_DEVICE_FLAGS_LOWER_UP_DORMANT_ECHO 1
+/* Definitions for if_ether.h */
+#define __UAPI_DEF_ETHHDR 1
+
/* Definitions for in.h */
#define __UAPI_DEF_IN_ADDR 1
#define __UAPI_DEF_IN_IPPROTO 1

40
target/linux/generic/pending-4.14/300-mips_expose_boot_raw.patch Normal file
View file

@ -0,0 +1,40 @@
From: Mark Miller <mark@mirell.org>
Subject: mips: expose CONFIG_BOOT_RAW
This exposes the CONFIG_BOOT_RAW symbol in Kconfig. This is needed on
certain Broadcom chipsets running CFE in order to load the kernel.
Signed-off-by: Mark Miller <mark@mirell.org>
Acked-by: Rob Landley <rob@landley.net>
---
--- a/arch/mips/Kconfig
+++ b/arch/mips/Kconfig
@@ -1066,9 +1066,6 @@ config FW_ARC
config ARCH_MAY_HAVE_PC_FDC
bool
-config BOOT_RAW
- bool
-
config CEVT_BCM1480
bool
@@ -2966,6 +2963,18 @@ choice
bool "Extend builtin kernel arguments with bootloader arguments"
endchoice
+config BOOT_RAW
+ bool "Enable the kernel to be executed from the load address"
+ default n
+ help
+ Allow the kernel to be executed from the load address for
+ bootloaders which cannot read the ELF format. This places
+ a jump to start_kernel at the load address.
+
+ If unsure, say N.
+
+
+
endmenu
config LOCKDEP_SUPPORT

22
target/linux/generic/pending-4.14/302-mips_no_branch_likely.patch Normal file
View file

@ -0,0 +1,22 @@
From: Felix Fietkau <nbd@nbd.name>
Subject: mips: use -mno-branch-likely for kernel and userspace
saves ~11k kernel size after lzma and ~12k squashfs size in the
lede-commit: 41a039f46450ffae9483d6216422098669da2900
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
arch/mips/Makefile | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
--- a/arch/mips/Makefile
+++ b/arch/mips/Makefile
@@ -90,7 +90,7 @@ all-$(CONFIG_SYS_SUPPORTS_ZBOOT)+= vmlin
# machines may also. Since BFD is incredibly buggy with respect to
# crossformat linking we rely on the elf2ecoff tool for format conversion.
#
-cflags-y += -G 0 -mno-abicalls -fno-pic -pipe
+cflags-y += -G 0 -mno-abicalls -fno-pic -pipe -mno-branch-likely
cflags-y += -msoft-float
LDFLAGS_vmlinux += -G 0 -static -n -nostdlib
KBUILD_AFLAGS_MODULE += -mlong-calls

137
target/linux/generic/pending-4.14/304-mips_disable_fpu.patch Normal file
View file

@ -0,0 +1,137 @@
From: Manuel Lauss <manuel.lauss@gmail.com>
Subject: [RFC PATCH v4 2/2] MIPS: make FPU emulator optional
This small patch makes the MIPS FPU emulator optional. The kernel
kills float-users on systems without a hardware FPU by sending a SIGILL.
Disabling the emulator shrinks vmlinux by about 54kBytes (32bit,
optimizing for size).
Signed-off-by: Manuel Lauss <manuel.lauss@gmail.com>
---
v4: rediffed because of patch 1/2, should now work with micromips as well
v3: updated patch description with size savings.
v2: incorporated changes suggested by Jonas Gorski
force the fpu emulator on for micromips: relocating the parts
of the mmips code in the emulator to other areas would be a
much larger change; I went the cheap route instead with this.
arch/mips/Kbuild | 2 +-
arch/mips/Kconfig | 14 ++++++++++++++
arch/mips/include/asm/fpu.h | 5 +++--
arch/mips/include/asm/fpu_emulator.h | 15 +++++++++++++++
4 files changed, 33 insertions(+), 3 deletions(-)
--- a/arch/mips/Kconfig
+++ b/arch/mips/Kconfig
@@ -2890,6 +2890,20 @@ config MIPS_O32_FP64_SUPPORT
If unsure, say N.
+config MIPS_FPU_EMULATOR
+ bool "MIPS FPU Emulator"
+ default y
+ help
+ This option lets you disable the built-in MIPS FPU (Coprocessor 1)
+ emulator, which handles floating-point instructions on processors
+ without a hardware FPU. It is generally a good idea to keep the
+ emulator built-in, unless you are perfectly sure you have a
+ complete soft-float environment. With the emulator disabled, all
+ users of float operations will be killed with an illegal instr-
+ uction exception.
+
+ Say Y, please.
+
config USE_OF
bool
select OF
--- a/arch/mips/Makefile
+++ b/arch/mips/Makefile
@@ -319,7 +319,7 @@ OBJCOPYFLAGS += --remove-section=.regin
head-y := arch/mips/kernel/head.o
libs-y += arch/mips/lib/
-libs-y += arch/mips/math-emu/
+libs-$(CONFIG_MIPS_FPU_EMULATOR) += arch/mips/math-emu/
# See arch/mips/Kbuild for content of core part of the kernel
core-y += arch/mips/
--- a/arch/mips/include/asm/fpu.h
+++ b/arch/mips/include/asm/fpu.h
@@ -230,8 +230,10 @@ static inline int init_fpu(void)
/* Restore FRE */
write_c0_config5(config5);
enable_fpu_hazard();
- } else
+ } else if (IS_ENABLED(CONFIG_MIPS_FPU_EMULATOR))
fpu_emulator_init_fpu();
+ else
+ ret = SIGILL;
return ret;
}
--- a/arch/mips/include/asm/fpu_emulator.h
+++ b/arch/mips/include/asm/fpu_emulator.h
@@ -30,6 +30,7 @@
#include <asm/local.h>
#include <asm/processor.h>
+#ifdef CONFIG_MIPS_FPU_EMULATOR
#ifdef CONFIG_DEBUG_FS
struct mips_fpu_emulator_stats {
@@ -179,6 +180,16 @@ do { \
extern int fpu_emulator_cop1Handler(struct pt_regs *xcp,
struct mips_fpu_struct *ctx, int has_fpu,
void __user **fault_addr);
+#else /* no CONFIG_MIPS_FPU_EMULATOR */
+static inline int fpu_emulator_cop1Handler(struct pt_regs *xcp,
+ struct mips_fpu_struct *ctx, int has_fpu,
+ void __user **fault_addr)
+{
+ *fault_addr = NULL;
+ return SIGILL; /* we don't speak MIPS FPU */
+}
+#endif /* CONFIG_MIPS_FPU_EMULATOR */
+
void force_fcr31_sig(unsigned long fcr31, void __user *fault_addr,
struct task_struct *tsk);
int process_fpemu_return(int sig, void __user *fault_addr,
--- a/arch/mips/include/asm/dsemul.h
+++ b/arch/mips/include/asm/dsemul.h
@@ -41,6 +41,7 @@ struct task_struct;
extern int mips_dsemul(struct pt_regs *regs, mips_instruction ir,
unsigned long branch_pc, unsigned long cont_pc);
+#ifdef CONFIG_MIPS_FPU_EMULATOR
/**
* do_dsemulret() - Return from a delay slot 'emulation' frame
* @xcp: User thread register context.
@@ -88,5 +89,27 @@ extern bool dsemul_thread_rollback(struc
* before @mm is freed in order to avoid memory leaks.
*/
extern void dsemul_mm_cleanup(struct mm_struct *mm);
+#else
+static inline bool do_dsemulret(struct pt_regs *xcp)
+{
+ return false;
+}
+
+static inline bool dsemul_thread_cleanup(struct task_struct *tsk)
+{
+ return false;
+}
+
+static inline bool dsemul_thread_rollback(struct pt_regs *regs)
+{
+ return false;
+}
+
+static inline void dsemul_mm_cleanup(struct mm_struct *mm)
+{
+
+}
+
+#endif
#endif /* __MIPS_ASM_DSEMUL_H__ */

366
target/linux/generic/pending-4.14/305-mips_module_reloc.patch Normal file
View file

@ -0,0 +1,366 @@
From: Felix Fietkau <nbd@nbd.name>
Subject: mips: replace -mlong-calls with -mno-long-calls to make function calls faster in kernel modules to achieve this, try to
lede-commit: 3b3d64743ba2a874df9d70cd19e242205b0a788c
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
arch/mips/Makefile | 5 +
arch/mips/include/asm/module.h | 5 +
arch/mips/kernel/module.c | 279 ++++++++++++++++++++++++++++++++++++++++-
3 files changed, 284 insertions(+), 5 deletions(-)
--- a/arch/mips/Makefile
+++ b/arch/mips/Makefile
@@ -93,8 +93,13 @@ all-$(CONFIG_SYS_SUPPORTS_ZBOOT)+= vmlin
cflags-y += -G 0 -mno-abicalls -fno-pic -pipe -mno-branch-likely
cflags-y += -msoft-float
LDFLAGS_vmlinux += -G 0 -static -n -nostdlib
+ifdef CONFIG_64BIT
KBUILD_AFLAGS_MODULE += -mlong-calls
KBUILD_CFLAGS_MODULE += -mlong-calls
+else
+KBUILD_AFLAGS_MODULE += -mno-long-calls
+KBUILD_CFLAGS_MODULE += -mno-long-calls
+endif
ifeq ($(CONFIG_RELOCATABLE),y)
LDFLAGS_vmlinux += --emit-relocs
--- a/arch/mips/include/asm/module.h
+++ b/arch/mips/include/asm/module.h
@@ -12,6 +12,11 @@ struct mod_arch_specific {
const struct exception_table_entry *dbe_start;
const struct exception_table_entry *dbe_end;
struct mips_hi16 *r_mips_hi16_list;
+
+ void *phys_plt_tbl;
+ void *virt_plt_tbl;
+ unsigned int phys_plt_offset;
+ unsigned int virt_plt_offset;
};
typedef uint8_t Elf64_Byte; /* Type for a 8-bit quantity. */
--- a/arch/mips/kernel/module.c
+++ b/arch/mips/kernel/module.c
@@ -44,14 +44,221 @@ struct mips_hi16 {
static LIST_HEAD(dbe_list);
static DEFINE_SPINLOCK(dbe_lock);
-#ifdef MODULE_START
+/*
+ * Get the potential max trampolines size required of the init and
+ * non-init sections. Only used if we cannot find enough contiguous
+ * physically mapped memory to put the module into.
+ */
+static unsigned int
+get_plt_size(const Elf_Ehdr *hdr, const Elf_Shdr *sechdrs,
+ const char *secstrings, unsigned int symindex, bool is_init)
+{
+ unsigned long ret = 0;
+ unsigned int i, j;
+ Elf_Sym *syms;
+
+ /* Everything marked ALLOC (this includes the exported symbols) */
+ for (i = 1; i < hdr->e_shnum; ++i) {
+ unsigned int info = sechdrs[i].sh_info;
+
+ if (sechdrs[i].sh_type != SHT_REL
+ && sechdrs[i].sh_type != SHT_RELA)
+ continue;
+
+ /* Not a valid relocation section? */
+ if (info >= hdr->e_shnum)
+ continue;
+
+ /* Don't bother with non-allocated sections */
+ if (!(sechdrs[info].sh_flags & SHF_ALLOC))
+ continue;
+
+ /* If it's called *.init*, and we're not init, we're
+ not interested */
+ if ((strstr(secstrings + sechdrs[i].sh_name, ".init") != 0)
+ != is_init)
+ continue;
+
+ syms = (Elf_Sym *) sechdrs[symindex].sh_addr;
+ if (sechdrs[i].sh_type == SHT_REL) {
+ Elf_Mips_Rel *rel = (void *) sechdrs[i].sh_addr;
+ unsigned int size = sechdrs[i].sh_size / sizeof(*rel);
+
+ for (j = 0; j < size; ++j) {
+ Elf_Sym *sym;
+
+ if (ELF_MIPS_R_TYPE(rel[j]) != R_MIPS_26)
+ continue;
+
+ sym = syms + ELF_MIPS_R_SYM(rel[j]);
+ if (!is_init && sym->st_shndx != SHN_UNDEF)
+ continue;
+
+ ret += 4 * sizeof(int);
+ }
+ } else {
+ Elf_Mips_Rela *rela = (void *) sechdrs[i].sh_addr;
+ unsigned int size = sechdrs[i].sh_size / sizeof(*rela);
+
+ for (j = 0; j < size; ++j) {
+ Elf_Sym *sym;
+
+ if (ELF_MIPS_R_TYPE(rela[j]) != R_MIPS_26)
+ continue;
+
+ sym = syms + ELF_MIPS_R_SYM(rela[j]);
+ if (!is_init && sym->st_shndx != SHN_UNDEF)
+ continue;
+
+ ret += 4 * sizeof(int);
+ }
+ }
+ }
+
+ return ret;
+}
+
+#ifndef MODULE_START
+static void *alloc_phys(unsigned long size)
+{
+ unsigned order;
+ struct page *page;
+ struct page *p;
+
+ size = PAGE_ALIGN(size);
+ order = get_order(size);
+
+ page = alloc_pages(GFP_KERNEL | __GFP_NORETRY | __GFP_NOWARN |
+ __GFP_THISNODE, order);
+ if (!page)
+ return NULL;
+
+ split_page(page, order);
+
+ /* mark all pages except for the last one */
+ for (p = page; p + 1 < page + (size >> PAGE_SHIFT); ++p)
+ set_bit(PG_owner_priv_1, &p->flags);
+
+ for (p = page + (size >> PAGE_SHIFT); p < page + (1 << order); ++p)
+ __free_page(p);
+
+ return page_address(page);
+}
+#endif
+
+static void free_phys(void *ptr)
+{
+ struct page *page;
+ bool free;
+
+ page = virt_to_page(ptr);
+ do {
+ free = test_and_clear_bit(PG_owner_priv_1, &page->flags);
+ __free_page(page);
+ page++;
+ } while (free);
+}
+
+
void *module_alloc(unsigned long size)
{
+#ifdef MODULE_START
return __vmalloc_node_range(size, 1, MODULE_START, MODULE_END,
GFP_KERNEL, PAGE_KERNEL, 0, NUMA_NO_NODE,
__builtin_return_address(0));
+#else
+ void *ptr;
+
+ if (size == 0)
+ return NULL;
+
+ ptr = alloc_phys(size);
+
+ /* If we failed to allocate physically contiguous memory,
+ * fall back to regular vmalloc. The module loader code will
+ * create jump tables to handle long jumps */
+ if (!ptr)
+ return vmalloc(size);
+
+ return ptr;
+#endif
}
+
+static inline bool is_phys_addr(void *ptr)
+{
+#ifdef CONFIG_64BIT
+ return (KSEGX((unsigned long)ptr) == CKSEG0);
+#else
+ return (KSEGX(ptr) == KSEG0);
#endif
+}
+
+/* Free memory returned from module_alloc */
+void module_memfree(void *module_region)
+{
+ if (is_phys_addr(module_region))
+ free_phys(module_region);
+ else
+ vfree(module_region);
+}
+
+static void *__module_alloc(int size, bool phys)
+{
+ void *ptr;
+
+ if (phys)
+ ptr = kmalloc(size, GFP_KERNEL);
+ else
+ ptr = vmalloc(size);
+ return ptr;
+}
+
+static void __module_free(void *ptr)
+{
+ if (is_phys_addr(ptr))
+ kfree(ptr);
+ else
+ vfree(ptr);
+}
+
+int module_frob_arch_sections(Elf_Ehdr *hdr, Elf_Shdr *sechdrs,
+ char *secstrings, struct module *mod)
+{
+ unsigned int symindex = 0;
+ unsigned int core_size, init_size;
+ int i;
+
+ mod->arch.phys_plt_offset = 0;
+ mod->arch.virt_plt_offset = 0;
+ mod->arch.phys_plt_tbl = NULL;
+ mod->arch.virt_plt_tbl = NULL;
+
+ if (IS_ENABLED(CONFIG_64BIT))
+ return 0;
+
+ for (i = 1; i < hdr->e_shnum; i++)
+ if (sechdrs[i].sh_type == SHT_SYMTAB)
+ symindex = i;
+
+ core_size = get_plt_size(hdr, sechdrs, secstrings, symindex, false);
+ init_size = get_plt_size(hdr, sechdrs, secstrings, symindex, true);
+
+ if ((core_size + init_size) == 0)
+ return 0;
+
+ mod->arch.phys_plt_tbl = __module_alloc(core_size + init_size, 1);
+ if (!mod->arch.phys_plt_tbl)
+ return -ENOMEM;
+
+ mod->arch.virt_plt_tbl = __module_alloc(core_size + init_size, 0);
+ if (!mod->arch.virt_plt_tbl) {
+ __module_free(mod->arch.phys_plt_tbl);
+ mod->arch.phys_plt_tbl = NULL;
+ return -ENOMEM;
+ }
+
+ return 0;
+}
static int apply_r_mips_none(struct module *me, u32 *location,
u32 base, Elf_Addr v, bool rela)
@@ -67,9 +274,40 @@ static int apply_r_mips_32(struct module
return 0;
}
+static Elf_Addr add_plt_entry_to(unsigned *plt_offset,
+ void *start, Elf_Addr v)
+{
+ unsigned *tramp = start + *plt_offset;
+ *plt_offset += 4 * sizeof(int);
+
+ /* adjust carry for addiu */
+ if (v & 0x00008000)
+ v += 0x10000;
+
+ tramp[0] = 0x3c190000 | (v >> 16); /* lui t9, hi16 */
+ tramp[1] = 0x27390000 | (v & 0xffff); /* addiu t9, t9, lo16 */
+ tramp[2] = 0x03200008; /* jr t9 */
+ tramp[3] = 0x00000000; /* nop */
+
+ return (Elf_Addr) tramp;
+}
+
+static Elf_Addr add_plt_entry(struct module *me, void *location, Elf_Addr v)
+{
+ if (is_phys_addr(location))
+ return add_plt_entry_to(&me->arch.phys_plt_offset,
+ me->arch.phys_plt_tbl, v);
+ else
+ return add_plt_entry_to(&me->arch.virt_plt_offset,
+ me->arch.virt_plt_tbl, v);
+
+}
+
static int apply_r_mips_26(struct module *me, u32 *location,
u32 base, Elf_Addr v, bool rela)
{
+ u32 ofs = base & 0x03ffffff;
+
if (v % 4) {
pr_err("module %s: dangerous R_MIPS_26 relocation\n",
me->name);
@@ -77,13 +315,17 @@ static int apply_r_mips_26(struct module
}
if ((v & 0xf0000000) != (((unsigned long)location + 4) & 0xf0000000)) {
- pr_err("module %s: relocation overflow\n",
- me->name);
- return -ENOEXEC;
+ v = add_plt_entry(me, location, v + (ofs << 2));
+ if (!v) {
+ pr_err("module %s: relocation overflow\n",
+ me->name);
+ return -ENOEXEC;
+ }
+ ofs = 0;
}
*location = (*location & ~0x03ffffff) |
- ((base + (v >> 2)) & 0x03ffffff);
+ ((ofs + (v >> 2)) & 0x03ffffff);
return 0;
}
@@ -459,9 +701,36 @@ int module_finalize(const Elf_Ehdr *hdr,
list_add(&me->arch.dbe_list, &dbe_list);
spin_unlock_irq(&dbe_lock);
}
+
+ /* Get rid of the fixup trampoline if we're running the module
+ * from physically mapped address space */
+ if (me->arch.phys_plt_offset == 0) {
+ __module_free(me->arch.phys_plt_tbl);
+ me->arch.phys_plt_tbl = NULL;
+ }
+ if (me->arch.virt_plt_offset == 0) {
+ __module_free(me->arch.virt_plt_tbl);
+ me->arch.virt_plt_tbl = NULL;
+ }
+
return 0;
}
+void module_arch_freeing_init(struct module *mod)
+{
+ if (mod->state == MODULE_STATE_LIVE)
+ return;
+
+ if (mod->arch.phys_plt_tbl) {
+ __module_free(mod->arch.phys_plt_tbl);
+ mod->arch.phys_plt_tbl = NULL;
+ }
+ if (mod->arch.virt_plt_tbl) {
+ __module_free(mod->arch.virt_plt_tbl);
+ mod->arch.virt_plt_tbl = NULL;
+ }
+}
+
void module_arch_cleanup(struct module *mod)
{
spin_lock_irq(&dbe_lock);

106
target/linux/generic/pending-4.14/306-mips_mem_functions_performance.patch Normal file
View file

@ -0,0 +1,106 @@
From: Felix Fietkau <nbd@nbd.name>
Subject: [PATCH] mips: allow the compiler to optimize memset, memcmp, memcpy for better performance and (in some instances) smaller code
lede-commit: 07e59c7bc7f375f792ec9734be42fe4fa391a8bb
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
arch/mips/boot/compressed/Makefile | 3 ++-
arch/mips/include/asm/string.h | 38 ++++++++++++++++++++++++++++++++++++++
arch/mips/lib/Makefile | 2 +-
arch/mips/lib/memcmp.c | 22 ++++++++++++++++++++++
4 files changed, 63 insertions(+), 2 deletions(-)
create mode 100644 arch/mips/lib/memcmp.c
--- a/arch/mips/boot/compressed/Makefile
+++ b/arch/mips/boot/compressed/Makefile
@@ -23,7 +23,8 @@ KBUILD_CFLAGS := $(filter-out -pg, $(KBU
KBUILD_CFLAGS := $(filter-out -fstack-protector, $(KBUILD_CFLAGS))
KBUILD_CFLAGS := $(KBUILD_CFLAGS) -D__KERNEL__ \
- -DBOOT_HEAP_SIZE=$(BOOT_HEAP_SIZE) -D"VMLINUX_LOAD_ADDRESS_ULL=$(VMLINUX_LOAD_ADDRESS)ull"
+ -DBOOT_HEAP_SIZE=$(BOOT_HEAP_SIZE) -D"VMLINUX_LOAD_ADDRESS_ULL=$(VMLINUX_LOAD_ADDRESS)ull" \
+ -D__ZBOOT__
KBUILD_AFLAGS := $(KBUILD_AFLAGS) -D__ASSEMBLY__ \
-DBOOT_HEAP_SIZE=$(BOOT_HEAP_SIZE) \
--- a/arch/mips/include/asm/string.h
+++ b/arch/mips/include/asm/string.h
@@ -140,4 +140,42 @@ extern void *memcpy(void *__to, __const_
#define __HAVE_ARCH_MEMMOVE
extern void *memmove(void *__dest, __const__ void *__src, size_t __n);
+#ifndef __ZBOOT__
+#define memset(__s, __c, len) \
+({ \
+ size_t __len = (len); \
+ void *__ret; \
+ if (__builtin_constant_p(len) && __len >= 64) \
+ __ret = memset((__s), (__c), __len); \
+ else \
+ __ret = __builtin_memset((__s), (__c), __len); \
+ __ret; \
+})
+
+#define memcpy(dst, src, len) \
+({ \
+ size_t __len = (len); \
+ void *__ret; \
+ if (__builtin_constant_p(len) && __len >= 64) \
+ __ret = memcpy((dst), (src), __len); \
+ else \
+ __ret = __builtin_memcpy((dst), (src), __len); \
+ __ret; \
+})
+
+#define memmove(dst, src, len) \
+({ \
+ size_t __len = (len); \
+ void *__ret; \
+ if (__builtin_constant_p(len) && __len >= 64) \
+ __ret = memmove((dst), (src), __len); \
+ else \
+ __ret = __builtin_memmove((dst), (src), __len); \
+ __ret; \
+})
+
+#define __HAVE_ARCH_MEMCMP
+#define memcmp(src1, src2, len) __builtin_memcmp((src1), (src2), (len))
+#endif
+
#endif /* _ASM_STRING_H */
--- a/arch/mips/lib/Makefile
+++ b/arch/mips/lib/Makefile
@@ -5,7 +5,7 @@
lib-y += bitops.o csum_partial.o delay.o memcpy.o memset.o \
mips-atomic.o strncpy_user.o \
- strnlen_user.o uncached.o
+ strnlen_user.o uncached.o memcmp.o
obj-y += iomap.o iomap_copy.o
obj-$(CONFIG_PCI) += iomap-pci.o
--- /dev/null
+++ b/arch/mips/lib/memcmp.c
@@ -0,0 +1,22 @@
+/*
+ * copied from linux/lib/string.c
+ *
+ * Copyright (C) 1991, 1992 Linus Torvalds
+ */
+
+#include <linux/module.h>
+#include <linux/string.h>
+
+#undef memcmp
+int memcmp(const void *cs, const void *ct, size_t count)
+{
+ const unsigned char *su1, *su2;
+ int res = 0;
+
+ for (su1 = cs, su2 = ct; 0 < count; ++su1, ++su2, count--)
+ if ((res = *su1 - *su2) != 0)
+ break;
+ return res;
+}
+EXPORT_SYMBOL(memcmp);
+

19
target/linux/generic/pending-4.14/307-mips_highmem_offset.patch Normal file
View file

@ -0,0 +1,19 @@
From: Felix Fietkau <nbd@nbd.name>
Subject: kernel: adjust mips highmem offset to avoid the need for -mlong-calls on systems with >256M RAM
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
arch/mips/include/asm/mach-generic/spaces.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
--- a/arch/mips/include/asm/mach-generic/spaces.h
+++ b/arch/mips/include/asm/mach-generic/spaces.h
@@ -46,7 +46,7 @@
* Memory above this physical address will be considered highmem.
*/
#ifndef HIGHMEM_START
-#define HIGHMEM_START _AC(0x20000000, UL)
+#define HIGHMEM_START _AC(0x10000000, UL)
#endif
#endif /* CONFIG_32BIT */

22
target/linux/generic/pending-4.14/308-mips32r2_tune.patch Normal file
View file

@ -0,0 +1,22 @@
From: Felix Fietkau <nbd@nbd.name>
Subject: kernel: add -mtune=34kc to MIPS CFLAGS when building for mips32r2
This provides a good tradeoff across at least 24Kc-74Kc, while also
producing smaller code.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
arch/mips/Makefile | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
--- a/arch/mips/Makefile
+++ b/arch/mips/Makefile
@@ -162,7 +162,7 @@ cflags-$(CONFIG_CPU_R4X00) += -march=r46
cflags-$(CONFIG_CPU_TX49XX) += -march=r4600 -Wa,--trap
cflags-$(CONFIG_CPU_MIPS32_R1) += $(call cc-option,-march=mips32,-mips32 -U_MIPS_ISA -D_MIPS_ISA=_MIPS_ISA_MIPS32) \
-Wa,-mips32 -Wa,--trap
-cflags-$(CONFIG_CPU_MIPS32_R2) += $(call cc-option,-march=mips32r2,-mips32r2 -U_MIPS_ISA -D_MIPS_ISA=_MIPS_ISA_MIPS32) \
+cflags-$(CONFIG_CPU_MIPS32_R2) += $(call cc-option,-march=mips32r2 -mtune=34kc,-mips32r2 -U_MIPS_ISA -D_MIPS_ISA=_MIPS_ISA_MIPS32) \
-Wa,-mips32r2 -Wa,--trap
cflags-$(CONFIG_CPU_MIPS32_R6) += -march=mips32r6 -Wa,--trap -modd-spreg
cflags-$(CONFIG_CPU_MIPS64_R1) += $(call cc-option,-march=mips64,-mips64 -U_MIPS_ISA -D_MIPS_ISA=_MIPS_ISA_MIPS64) \

22
target/linux/generic/pending-4.14/310-arm_module_unresolved_weak_sym.patch Normal file
View file

@ -0,0 +1,22 @@
From: Felix Fietkau <nbd@nbd.name>
Subject: fix errors in unresolved weak symbols on arm
lede-commit: 570699d4838a907c3ef9f2819bf19eb72997b32f
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
arch/arm/kernel/module.c | 4 ++++
1 file changed, 4 insertions(+)
--- a/arch/arm/kernel/module.c
+++ b/arch/arm/kernel/module.c
@@ -95,6 +95,10 @@ apply_relocate(Elf32_Shdr *sechdrs, cons
return -ENOEXEC;
}
+ if ((IS_ERR_VALUE(sym->st_value) || !sym->st_value) &&
+ ELF_ST_BIND(sym->st_info) == STB_WEAK)
+ continue;
+
loc = dstsec->sh_addr + rel->r_offset;
switch (ELF32_R_TYPE(rel->r_info)) {

272
target/linux/generic/pending-4.14/330-MIPS-kexec-Accept-command-line-parameters-from-users.patch Normal file
View file

@ -0,0 +1,272 @@
From: Yousong Zhou <yszhou4tech@gmail.com>
Subject: MIPS: kexec: Accept command line parameters from userspace.
Signed-off-by: Yousong Zhou <yszhou4tech@gmail.com>
---
arch/mips/kernel/machine_kexec.c | 153 +++++++++++++++++++++++++++++++-----
arch/mips/kernel/machine_kexec.h | 20 +++++
arch/mips/kernel/relocate_kernel.S | 21 +++--
3 files changed, 167 insertions(+), 27 deletions(-)
create mode 100644 arch/mips/kernel/machine_kexec.h
--- a/arch/mips/kernel/machine_kexec.c
+++ b/arch/mips/kernel/machine_kexec.c
@@ -10,14 +10,11 @@
#include <linux/mm.h>
#include <linux/delay.h>
+#include <asm/bootinfo.h>
#include <asm/cacheflush.h>
#include <asm/page.h>
-
-extern const unsigned char relocate_new_kernel[];
-extern const size_t relocate_new_kernel_size;
-
-extern unsigned long kexec_start_address;
-extern unsigned long kexec_indirection_page;
+#include <asm/uaccess.h>
+#include "machine_kexec.h"
int (*_machine_kexec_prepare)(struct kimage *) = NULL;
void (*_machine_kexec_shutdown)(void) = NULL;
@@ -28,6 +25,99 @@ atomic_t kexec_ready_to_reboot = ATOMIC_
void (*_crash_smp_send_stop)(void) = NULL;
#endif
+static void machine_kexec_print_args(void)
+{
+ unsigned long argc = (int)kexec_args[0];
+ int i;
+
+ pr_info("kexec_args[0] (argc): %lu\n", argc);
+ pr_info("kexec_args[1] (argv): %p\n", (void *)kexec_args[1]);
+ pr_info("kexec_args[2] (env ): %p\n", (void *)kexec_args[2]);
+ pr_info("kexec_args[3] (desc): %p\n", (void *)kexec_args[3]);
+
+ for (i = 0; i < argc; i++) {
+ pr_info("kexec_argv[%d] = %p, %s\n",
+ i, kexec_argv[i], kexec_argv[i]);
+ }
+}
+
+static void machine_kexec_init_argv(struct kimage *image)
+{
+ void __user *buf = NULL;
+ size_t bufsz;
+ size_t size;
+ int i;
+
+ bufsz = 0;
+ for (i = 0; i < image->nr_segments; i++) {
+ struct kexec_segment *seg;
+
+ seg = &image->segment[i];
+ if (seg->bufsz < 6)
+ continue;
+
+ if (strncmp((char *) seg->buf, "kexec ", 6))
+ continue;
+
+ buf = seg->buf;
+ bufsz = seg->bufsz;
+ break;
+ }
+
+ if (!buf)
+ return;
+
+ size = KEXEC_COMMAND_LINE_SIZE;
+ size = min(size, bufsz);
+ if (size < bufsz)
+ pr_warn("kexec command line truncated to %zd bytes\n", size);
+
+ /* Copy to kernel space */
+ copy_from_user(kexec_argv_buf, buf, size);
+ kexec_argv_buf[size - 1] = 0;
+}
+
+static void machine_kexec_parse_argv(struct kimage *image)
+{
+ char *reboot_code_buffer;
+ int reloc_delta;
+ char *ptr;
+ int argc;
+ int i;
+
+ ptr = kexec_argv_buf;
+ argc = 0;
+
+ /*
+ * convert command line string to array of parameters
+ * (as bootloader does).
+ */
+ while (ptr && *ptr && (KEXEC_MAX_ARGC > argc)) {
+ if (*ptr == ' ') {
+ *ptr++ = '\0';
+ continue;
+ }
+
+ kexec_argv[argc++] = ptr;
+ ptr = strchr(ptr, ' ');
+ }
+
+ if (!argc)
+ return;
+
+ kexec_args[0] = argc;
+ kexec_args[1] = (unsigned long)kexec_argv;
+ kexec_args[2] = 0;
+ kexec_args[3] = 0;
+
+ reboot_code_buffer = page_address(image->control_code_page);
+ reloc_delta = reboot_code_buffer - (char *)kexec_relocate_new_kernel;
+
+ kexec_args[1] += reloc_delta;
+ for (i = 0; i < argc; i++)
+ kexec_argv[i] += reloc_delta;
+}
+
static void kexec_image_info(const struct kimage *kimage)
{
unsigned long i;
@@ -52,6 +142,18 @@ int
machine_kexec_prepare(struct kimage *kimage)
{
kexec_image_info(kimage);
+ /*
+ * Whenever arguments passed from kexec-tools, Init the arguments as
+ * the original ones to try avoiding booting failure.
+ */
+
+ kexec_args[0] = fw_arg0;
+ kexec_args[1] = fw_arg1;
+ kexec_args[2] = fw_arg2;
+ kexec_args[3] = fw_arg3;
+
+ machine_kexec_init_argv(kimage);
+ machine_kexec_parse_argv(kimage);
if (_machine_kexec_prepare)
return _machine_kexec_prepare(kimage);
@@ -89,10 +191,12 @@ machine_kexec(struct kimage *image)
unsigned long *ptr;
reboot_code_buffer =
- (unsigned long)page_address(image->control_code_page);
+ (unsigned long)page_address(image->control_code_page);
+ pr_info("reboot_code_buffer = %p\n", (void *)reboot_code_buffer);
kexec_start_address =
(unsigned long) phys_to_virt(image->start);
+ pr_info("kexec_start_address = %p\n", (void *)kexec_start_address);
if (image->type == KEXEC_TYPE_DEFAULT) {
kexec_indirection_page =
@@ -100,9 +204,19 @@ machine_kexec(struct kimage *image)
} else {
kexec_indirection_page = (unsigned long)&image->head;
}
+ pr_info("kexec_indirection_page = %p\n", (void *)kexec_indirection_page);
- memcpy((void*)reboot_code_buffer, relocate_new_kernel,
- relocate_new_kernel_size);
+ pr_info("Where is memcpy: %p\n", memcpy);
+ pr_info("kexec_relocate_new_kernel = %p, kexec_relocate_new_kernel_end = %p\n",
+ (void *)kexec_relocate_new_kernel, &kexec_relocate_new_kernel_end);
+ pr_info("Copy %lu bytes from %p to %p\n", KEXEC_RELOCATE_NEW_KERNEL_SIZE,
+ (void *)kexec_relocate_new_kernel, (void *)reboot_code_buffer);
+ memcpy((void*)reboot_code_buffer, kexec_relocate_new_kernel,
+ KEXEC_RELOCATE_NEW_KERNEL_SIZE);
+
+ pr_info("Before _print_args().\n");
+ machine_kexec_print_args();
+ pr_info("Before eval loop.\n");
/*
* The generic kexec code builds a page list with physical
@@ -121,15 +235,16 @@ machine_kexec(struct kimage *image)
/*
* we do not want to be bothered.
*/
+ pr_info("Before irq_disable.\n");
local_irq_disable();
- printk("Will call new kernel at %08lx\n", image->start);
- printk("Bye ...\n");
+ pr_info("Will call new kernel at %08lx\n", image->start);
+ pr_info("Bye ...\n");
__flush_cache_all();
#ifdef CONFIG_SMP
/* All secondary cpus now may jump to kexec_wait cycle */
relocated_kexec_smp_wait = reboot_code_buffer +
- (void *)(kexec_smp_wait - relocate_new_kernel);
+ (void *)(kexec_smp_wait - kexec_relocate_new_kernel);
smp_wmb();
atomic_set(&kexec_ready_to_reboot, 1);
#endif
--- /dev/null
+++ b/arch/mips/kernel/machine_kexec.h
@@ -0,0 +1,20 @@
+#ifndef _MACHINE_KEXEC_H
+#define _MACHINE_KEXEC_H
+
+#ifndef __ASSEMBLY__
+extern const unsigned char kexec_relocate_new_kernel[];
+extern unsigned long kexec_relocate_new_kernel_end;
+extern unsigned long kexec_start_address;
+extern unsigned long kexec_indirection_page;
+
+extern char kexec_argv_buf[];
+extern char *kexec_argv[];
+
+#define KEXEC_RELOCATE_NEW_KERNEL_SIZE ((unsigned long)&kexec_relocate_new_kernel_end - (unsigned long)kexec_relocate_new_kernel)
+#endif /* !__ASSEMBLY__ */
+
+#define KEXEC_COMMAND_LINE_SIZE 256
+#define KEXEC_ARGV_SIZE (KEXEC_COMMAND_LINE_SIZE / 16)
+#define KEXEC_MAX_ARGC (KEXEC_ARGV_SIZE / sizeof(long))
+
+#endif
--- a/arch/mips/kernel/relocate_kernel.S
+++ b/arch/mips/kernel/relocate_kernel.S
@@ -12,8 +12,9 @@
#include <asm/mipsregs.h>
#include <asm/stackframe.h>
#include <asm/addrspace.h>
+#include "machine_kexec.h"
-LEAF(relocate_new_kernel)
+LEAF(kexec_relocate_new_kernel)
PTR_L a0, arg0
PTR_L a1, arg1
PTR_L a2, arg2
@@ -98,7 +99,7 @@ done:
#endif
/* jump to kexec_start_address */
j s1
- END(relocate_new_kernel)
+ END(kexec_relocate_new_kernel)
#ifdef CONFIG_SMP
/*
@@ -184,9 +185,15 @@ kexec_indirection_page:
PTR 0
.size kexec_indirection_page, PTRSIZE
-relocate_new_kernel_end:
+kexec_argv_buf:
+ EXPORT(kexec_argv_buf)
+ .skip KEXEC_COMMAND_LINE_SIZE
+ .size kexec_argv_buf, KEXEC_COMMAND_LINE_SIZE
+
+kexec_argv:
+ EXPORT(kexec_argv)
+ .skip KEXEC_ARGV_SIZE
+ .size kexec_argv, KEXEC_ARGV_SIZE
-relocate_new_kernel_size:
- EXPORT(relocate_new_kernel_size)
- PTR relocate_new_kernel_end - relocate_new_kernel
- .size relocate_new_kernel_size, PTRSIZE
+kexec_relocate_new_kernel_end:
+ EXPORT(kexec_relocate_new_kernel_end)

80
target/linux/generic/pending-4.14/332-arc-add-OWRTDTB-section.patch Normal file
View file

@ -0,0 +1,80 @@
From: Alexey Brodkin <abrodkin@synopsys.com>
Subject: openwrt: arc - add OWRTDTB section
This change allows OpenWRT to patch resulting kernel binary with
external .dtb.
That allows us to re-use exactky the same vmlinux on different boards
given its ARC core configurations match (at least cache line sizes etc).
""patch-dtb" searches for ASCII "OWRTDTB:" strign and copies external
.dtb right after it, keeping the string in place.
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
---
arch/arc/kernel/head.S | 10 ++++++++++
arch/arc/kernel/setup.c | 4 +++-
arch/arc/kernel/vmlinux.lds.S | 13 +++++++++++++
3 files changed, 26 insertions(+), 1 deletion(-)
--- a/arch/arc/kernel/head.S
+++ b/arch/arc/kernel/head.S
@@ -49,6 +49,16 @@
1:
.endm
+; Here "patch-dtb" will embed external .dtb
+; Note "patch-dtb" searches for ASCII "OWRTDTB:" string
+; and pastes .dtb right after it, hense the string precedes
+; __image_dtb symbol.
+ .section .owrt, "aw",@progbits
+ .ascii "OWRTDTB:"
+ENTRY(__image_dtb)
+ .fill 0x4000
+END(__image_dtb)
+
.section .init.text, "ax",@progbits
;----------------------------------------------------------------
--- a/arch/arc/kernel/setup.c
+++ b/arch/arc/kernel/setup.c
@@ -421,6 +421,8 @@ static inline int is_kernel(unsigned lon
return 0;
}
+extern struct boot_param_header __image_dtb;
+
void __init setup_arch(char **cmdline_p)
{
#ifdef CONFIG_ARC_UBOOT_SUPPORT
@@ -434,7 +436,7 @@ void __init setup_arch(char **cmdline_p)
#endif
{
/* No, so try the embedded one */
- machine_desc = setup_machine_fdt(__dtb_start);
+ machine_desc = setup_machine_fdt(&__image_dtb);
if (!machine_desc)
panic("Embedded DT invalid\n");
--- a/arch/arc/kernel/vmlinux.lds.S
+++ b/arch/arc/kernel/vmlinux.lds.S
@@ -30,6 +30,19 @@ SECTIONS
. = CONFIG_LINUX_LINK_BASE;
+ /*
+ * In OpenWRT we want to patch built binary embedding .dtb of choice.
+ * This is implemented with "patch-dtb" utility which searches for
+ * "OWRTDTB:" string in first 16k of image and if it is found
+ * copies .dtb right after mentioned string.
+ *
+ * Note: "OWRTDTB:" won't be overwritten with .dtb, .dtb will follow it.
+ */
+ .owrt : {
+ *(.owrt)
+ . = ALIGN(PAGE_SIZE);
+ }
+
_int_vec_base_lds = .;
.vector : {
*(.vector)

24
target/linux/generic/pending-4.14/333-arc-enable-unaligned-access-in-kernel-mode.patch Normal file
View file

@ -0,0 +1,24 @@
From: Alexey Brodkin <abrodkin@synopsys.com>
Subject: arc: enable unaligned access in kernel mode
This enables misaligned access handling even in kernel mode.
Some wireless drivers (ath9k-htc and mt7601u) use misaligned accesses
here and there and to cope with that without fixing stuff in the drivers
we're just gracefully handling it on ARC.
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
---
arch/arc/kernel/unaligned.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
--- a/arch/arc/kernel/unaligned.c
+++ b/arch/arc/kernel/unaligned.c
@@ -206,7 +206,7 @@ int misaligned_fixup(unsigned long addre
char buf[TASK_COMM_LEN];
/* handle user mode only and only if enabled by sysadmin */
- if (!user_mode(regs) || !unaligned_enabled)
+ if (!unaligned_enabled)
return 1;
if (no_unaligned_warning) {

32
target/linux/generic/pending-4.14/340-MIPS-mm-remove-mips_dma_mapping_error.patch Normal file
View file

@ -0,0 +1,32 @@
From: Felix Fietkau <nbd@nbd.name>
Date: Tue, 5 Dec 2017 12:34:31 +0100
Subject: [PATCH] MIPS: mm: remove mips_dma_mapping_error
dma_mapping_error() already checks if ops->mapping_error is a null
pointer
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
--- a/arch/mips/mm/dma-default.c
+++ b/arch/mips/mm/dma-default.c
@@ -373,11 +373,6 @@ static void mips_dma_sync_sg_for_device(
}
}
-static int mips_dma_mapping_error(struct device *dev, dma_addr_t dma_addr)
-{
- return 0;
-}
-
static int mips_dma_supported(struct device *dev, u64 mask)
{
return plat_dma_supported(dev, mask);
@@ -406,7 +401,6 @@ static const struct dma_map_ops mips_def
.sync_single_for_device = mips_dma_sync_single_for_device,
.sync_sg_for_cpu = mips_dma_sync_sg_for_cpu,
.sync_sg_for_device = mips_dma_sync_sg_for_device,
- .mapping_error = mips_dma_mapping_error,
.dma_supported = mips_dma_supported
};

140
target/linux/generic/pending-4.14/341-MIPS-mm-remove-no-op-dma_map_ops-where-possible.patch Normal file
View file

@ -0,0 +1,140 @@
From: Felix Fietkau <nbd@nbd.name>
Date: Tue, 5 Dec 2017 12:46:01 +0100
Subject: [PATCH] MIPS: mm: remove no-op dma_map_ops where possible
If no post-DMA flush is required, and the platform does not provide
plat_unmap_dma_mem(), there is no need to include unmap or sync_for_cpu
ops.
With this patch they are compiled out to improve icache footprint
on devices that handle lots of DMA traffic (especially network routers).
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
--- a/arch/mips/Kconfig
+++ b/arch/mips/Kconfig
@@ -220,6 +220,7 @@ config BMIPS_GENERIC
select BRCMSTB_L2_IRQ
select IRQ_MIPS_CPU
select DMA_NONCOHERENT
+ select DMA_UNMAP_POST_FLUSH
select SYS_SUPPORTS_32BIT_KERNEL
select SYS_SUPPORTS_LITTLE_ENDIAN
select SYS_SUPPORTS_BIG_ENDIAN
@@ -345,6 +346,7 @@ config MACH_JAZZ
select CSRC_R4K
select DEFAULT_SGI_PARTITION if CPU_BIG_ENDIAN
select GENERIC_ISA_DMA
+ select DMA_UNMAP_POST_FLUSH
select HAVE_PCSPKR_PLATFORM
select IRQ_MIPS_CPU
select I8253
@@ -1127,6 +1129,9 @@ config DMA_NONCOHERENT
bool
select NEED_DMA_MAP_STATE
+config DMA_UNMAP_POST_FLUSH
+ bool
+
config NEED_DMA_MAP_STATE
bool
@@ -1651,6 +1656,7 @@ config CPU_R10000
select CPU_SUPPORTS_64BIT_KERNEL
select CPU_SUPPORTS_HIGHMEM
select CPU_SUPPORTS_HUGEPAGES
+ select DMA_UNMAP_POST_FLUSH
help
MIPS Technologies R10000-series processors.
@@ -1899,9 +1905,11 @@ config SYS_HAS_CPU_MIPS32_R3_5
bool
config SYS_HAS_CPU_MIPS32_R5
+ select DMA_UNMAP_POST_FLUSH
bool
config SYS_HAS_CPU_MIPS32_R6
+ select DMA_UNMAP_POST_FLUSH
bool
config SYS_HAS_CPU_MIPS64_R1
@@ -1911,6 +1919,7 @@ config SYS_HAS_CPU_MIPS64_R2
bool
config SYS_HAS_CPU_MIPS64_R6
+ select DMA_UNMAP_POST_FLUSH
bool
config SYS_HAS_CPU_R3000
--- a/arch/mips/mm/dma-default.c
+++ b/arch/mips/mm/dma-default.c
@@ -267,8 +267,9 @@ static inline void __dma_sync(struct pag
} while (left);
}
-static void mips_dma_unmap_page(struct device *dev, dma_addr_t dma_addr,
- size_t size, enum dma_data_direction direction, unsigned long attrs)
+static void __maybe_unused
+mips_dma_unmap_page(struct device *dev, dma_addr_t dma_addr, size_t size,
+ enum dma_data_direction direction, unsigned long attrs)
{
if (cpu_needs_post_dma_flush(dev) && !(attrs & DMA_ATTR_SKIP_CPU_SYNC))
__dma_sync(dma_addr_to_page(dev, dma_addr),
@@ -308,9 +309,10 @@ static dma_addr_t mips_dma_map_page(stru
return plat_map_dma_mem_page(dev, page) + offset;
}
-static void mips_dma_unmap_sg(struct device *dev, struct scatterlist *sglist,
- int nhwentries, enum dma_data_direction direction,
- unsigned long attrs)
+static void __maybe_unused
+mips_dma_unmap_sg(struct device *dev, struct scatterlist *sglist,
+ int nhwentries, enum dma_data_direction direction,
+ unsigned long attrs)
{
int i;
struct scatterlist *sg;
@@ -325,8 +327,9 @@ static void mips_dma_unmap_sg(struct dev
}
}
-static void mips_dma_sync_single_for_cpu(struct device *dev,
- dma_addr_t dma_handle, size_t size, enum dma_data_direction direction)
+static void __maybe_unused
+mips_dma_sync_single_for_cpu(struct device *dev, dma_addr_t dma_handle,
+ size_t size, enum dma_data_direction direction)
{
if (cpu_needs_post_dma_flush(dev))
__dma_sync(dma_addr_to_page(dev, dma_handle),
@@ -342,9 +345,9 @@ static void mips_dma_sync_single_for_dev
dma_handle & ~PAGE_MASK, size, direction);
}
-static void mips_dma_sync_sg_for_cpu(struct device *dev,
- struct scatterlist *sglist, int nelems,
- enum dma_data_direction direction)
+static void __maybe_unused
+mips_dma_sync_sg_for_cpu(struct device *dev, struct scatterlist *sglist,
+ int nelems, enum dma_data_direction direction)
{
int i;
struct scatterlist *sg;
@@ -394,12 +397,14 @@ static const struct dma_map_ops mips_def
.free = mips_dma_free_coherent,
.mmap = mips_dma_mmap,
.map_page = mips_dma_map_page,
- .unmap_page = mips_dma_unmap_page,
.map_sg = mips_dma_map_sg,
+#ifdef CONFIG_DMA_UNMAP_POST_FLUSH
+ .unmap_page = mips_dma_unmap_page,
.unmap_sg = mips_dma_unmap_sg,
.sync_single_for_cpu = mips_dma_sync_single_for_cpu,
- .sync_single_for_device = mips_dma_sync_single_for_device,
.sync_sg_for_cpu = mips_dma_sync_sg_for_cpu,
+#endif
+ .sync_single_for_device = mips_dma_sync_single_for_device,
.sync_sg_for_device = mips_dma_sync_sg_for_device,
.dma_supported = mips_dma_supported
};

123
target/linux/generic/pending-4.14/400-mtd-add-rootfs-split-support.patch Normal file
View file

@ -0,0 +1,123 @@
From: Felix Fietkau <nbd@nbd.name>
Subject: make rootfs split/detection more generic - patch can be moved to generic-2.6 after testing on other platforms
lede-commit: 328e660b31f0937d52c5ae3d6e7029409918a9df
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
drivers/mtd/Kconfig | 17 +++++++++++++++++
drivers/mtd/mtdpart.c | 35 +++++++++++++++++++++++++++++++++++
include/linux/mtd/partitions.h | 2 ++
3 files changed, 54 insertions(+)
--- a/drivers/mtd/Kconfig
+++ b/drivers/mtd/Kconfig
@@ -12,6 +12,23 @@ menuconfig MTD
if MTD
+menu "OpenWrt specific MTD options"
+
+config MTD_ROOTFS_ROOT_DEV
+ bool "Automatically set 'rootfs' partition to be root filesystem"
+ default y
+
+config MTD_SPLIT_FIRMWARE
+ bool "Automatically split firmware partition for kernel+rootfs"
+ default y
+
+config MTD_SPLIT_FIRMWARE_NAME
+ string "Firmware partition name"
+ depends on MTD_SPLIT_FIRMWARE
+ default "firmware"
+
+endmenu
+
config MTD_TESTS
tristate "MTD tests support (DANGEROUS)"
depends on m
--- a/drivers/mtd/mtdpart.c
+++ b/drivers/mtd/mtdpart.c
@@ -29,10 +29,12 @@
#include <linux/kmod.h>
#include <linux/mtd/mtd.h>
#include <linux/mtd/partitions.h>
+#include <linux/magic.h>
#include <linux/of.h>
#include <linux/err.h>
#include "mtdcore.h"
+#include "mtdsplit/mtdsplit.h"
/* Our partition linked list */
static LIST_HEAD(mtd_partitions);
@@ -52,6 +54,8 @@ struct mtd_part {
struct list_head list;
};
+static void mtd_partition_split(struct mtd_info *master, struct mtd_part *part);
+
/*
* Given a pointer to the MTD object in the mtd_part structure, we can retrieve
* the pointer to that structure.
@@ -686,6 +690,7 @@ int mtd_add_partition(struct mtd_info *p
mutex_unlock(&mtd_partitions_mutex);
add_mtd_device(&new->mtd);
+ mtd_partition_split(parent, new);
mtd_add_partition_attrs(new);
@@ -764,6 +769,35 @@ int mtd_del_partition(struct mtd_info *m
}
EXPORT_SYMBOL_GPL(mtd_del_partition);
+#ifdef CONFIG_MTD_SPLIT_FIRMWARE_NAME
+#define SPLIT_FIRMWARE_NAME CONFIG_MTD_SPLIT_FIRMWARE_NAME
+#else
+#define SPLIT_FIRMWARE_NAME "unused"
+#endif
+
+static void split_firmware(struct mtd_info *master, struct mtd_part *part)
+{
+}
+
+void __weak arch_split_mtd_part(struct mtd_info *master, const char *name,
+ int offset, int size)
+{
+}
+
+static void mtd_partition_split(struct mtd_info *master, struct mtd_part *part)
+{
+ static int rootfs_found = 0;
+
+ if (rootfs_found)
+ return;
+
+ if (!strcmp(part->mtd.name, SPLIT_FIRMWARE_NAME) &&
+ IS_ENABLED(CONFIG_MTD_SPLIT_FIRMWARE))
+ split_firmware(master, part);
+
+ arch_split_mtd_part(master, part->mtd.name, part->offset,
+ part->mtd.size);
+}
/*
* This function, given a master MTD object and a partition table, creates
* and registers slave MTD objects which are bound to the master according to
@@ -795,6 +829,7 @@ int add_mtd_partitions(struct mtd_info *
mutex_unlock(&mtd_partitions_mutex);
add_mtd_device(&slave->mtd);
+ mtd_partition_split(master, slave);
mtd_add_partition_attrs(slave);
if (parts[i].types)
mtd_parse_part(slave, parts[i].types);
--- a/include/linux/mtd/partitions.h
+++ b/include/linux/mtd/partitions.h
@@ -109,5 +109,7 @@ int mtd_add_partition(struct mtd_info *m
long long offset, long long length);
int mtd_del_partition(struct mtd_info *master, int partno);
uint64_t mtd_get_device_size(const struct mtd_info *mtd);
+extern void __weak arch_split_mtd_part(struct mtd_info *master,
+ const char *name, int offset, int size);
#endif

110
target/linux/generic/pending-4.14/401-mtd-add-support-for-different-partition-parser-types.patch Normal file
View file

@ -0,0 +1,110 @@
From: Gabor Juhos <juhosg@openwrt.org>
Subject: mtd: add support for different partition parser types
Signed-off-by: Gabor Juhos <juhosg@openwrt.org>
---
drivers/mtd/mtdpart.c | 56 ++++++++++++++++++++++++++++++++++++++++
include/linux/mtd/partitions.h | 11 ++++++++
2 files changed, 67 insertions(+)
--- a/drivers/mtd/mtdpart.c
+++ b/drivers/mtd/mtdpart.c
@@ -1034,6 +1034,62 @@ void mtd_part_parser_cleanup(struct mtd_
}
}
+static struct mtd_part_parser *
+get_partition_parser_by_type(enum mtd_parser_type type,
+ struct mtd_part_parser *start)
+{
+ struct mtd_part_parser *p, *ret = NULL;
+
+ spin_lock(&part_parser_lock);
+
+ p = list_prepare_entry(start, &part_parsers, list);
+ if (start)
+ mtd_part_parser_put(start);
+
+ list_for_each_entry_continue(p, &part_parsers, list) {
+ if (p->type == type && try_module_get(p->owner)) {
+ ret = p;
+ break;
+ }
+ }
+
+ spin_unlock(&part_parser_lock);
+
+ return ret;
+}
+
+int parse_mtd_partitions_by_type(struct mtd_info *master,
+ enum mtd_parser_type type,
+ const struct mtd_partition **pparts,
+ struct mtd_part_parser_data *data)
+{
+ struct mtd_part_parser *prev = NULL;
+ int ret = 0;
+
+ while (1) {
+ struct mtd_part_parser *parser;
+
+ parser = get_partition_parser_by_type(type, prev);
+ if (!parser)
+ break;
+
+ ret = (*parser->parse_fn)(master, pparts, data);
+
+ if (ret > 0) {
+ mtd_part_parser_put(parser);
+ printk(KERN_NOTICE
+ "%d %s partitions found on MTD device %s\n",
+ ret, parser->name, master->name);
+ break;
+ }
+
+ prev = parser;
+ }
+
+ return ret;
+}
+EXPORT_SYMBOL_GPL(parse_mtd_partitions_by_type);
+
int mtd_is_partition(const struct mtd_info *mtd)
{
struct mtd_part *part;
--- a/include/linux/mtd/partitions.h
+++ b/include/linux/mtd/partitions.h
@@ -68,11 +68,14 @@ struct mtd_part_parser_data {
unsigned long origin;
};
-
/*
* Functions dealing with the various ways of partitioning the space
*/
+enum mtd_parser_type {
+ MTD_PARSER_TYPE_DEVICE = 0,
+};
+
struct mtd_part_parser {
struct list_head list;
struct module *owner;
@@ -80,6 +83,7 @@ struct mtd_part_parser {
int (*parse_fn)(struct mtd_info *, const struct mtd_partition **,
struct mtd_part_parser_data *);
void (*cleanup)(const struct mtd_partition *pparts, int nr_parts);
+ enum mtd_parser_type type;
};
/* Container for passing around a set of parsed partitions */
@@ -112,4 +116,9 @@ uint64_t mtd_get_device_size(const struc
extern void __weak arch_split_mtd_part(struct mtd_info *master,
const char *name, int offset, int size);
+int parse_mtd_partitions_by_type(struct mtd_info *master,
+ enum mtd_parser_type type,
+ const struct mtd_partition **pparts,
+ struct mtd_part_parser_data *data);
+
#endif

81
target/linux/generic/pending-4.14/402-mtd-use-typed-mtd-parsers-for-rootfs-and-firmware-split.patch Normal file
View file

@ -0,0 +1,81 @@
From: Gabor Juhos <juhosg@openwrt.org>
Subject: kernel/3.10: allow to use partition parsers for rootfs and firmware split
lede-commit: 3b71cd94bc9517bc25267dccb393b07d4b54564e
Signed-off-by: Gabor Juhos <juhosg@openwrt.org>
---
drivers/mtd/mtdpart.c | 37 +++++++++++++++++++++++++++++++++++++
include/linux/mtd/partitions.h | 2 ++
2 files changed, 39 insertions(+)
--- a/drivers/mtd/mtdpart.c
+++ b/drivers/mtd/mtdpart.c
@@ -769,6 +769,36 @@ int mtd_del_partition(struct mtd_info *m
}
EXPORT_SYMBOL_GPL(mtd_del_partition);
+static int
+run_parsers_by_type(struct mtd_part *slave, enum mtd_parser_type type)
+{
+ struct mtd_partition *parts;
+ int nr_parts;
+ int i;
+
+ nr_parts = parse_mtd_partitions_by_type(&slave->mtd, type, (const struct mtd_partition **)&parts,
+ NULL);
+ if (nr_parts <= 0)
+ return nr_parts;
+
+ if (WARN_ON(!parts))
+ return 0;
+
+ for (i = 0; i < nr_parts; i++) {
+ /* adjust partition offsets */
+ parts[i].offset += slave->offset;
+
+ mtd_add_partition(slave->parent,
+ parts[i].name,
+ parts[i].offset,
+ parts[i].size);
+ }
+
+ kfree(parts);
+
+ return nr_parts;
+}
+
#ifdef CONFIG_MTD_SPLIT_FIRMWARE_NAME
#define SPLIT_FIRMWARE_NAME CONFIG_MTD_SPLIT_FIRMWARE_NAME
#else
@@ -777,6 +807,7 @@ EXPORT_SYMBOL_GPL(mtd_del_partition);
static void split_firmware(struct mtd_info *master, struct mtd_part *part)
{
+ run_parsers_by_type(part, MTD_PARSER_TYPE_FIRMWARE);
}
void __weak arch_split_mtd_part(struct mtd_info *master, const char *name,
@@ -791,6 +822,12 @@ static void mtd_partition_split(struct m
if (rootfs_found)
return;
+ if (!strcmp(part->mtd.name, "rootfs")) {
+ run_parsers_by_type(part, MTD_PARSER_TYPE_ROOTFS);
+
+ rootfs_found = 1;
+ }
+
if (!strcmp(part->mtd.name, SPLIT_FIRMWARE_NAME) &&
IS_ENABLED(CONFIG_MTD_SPLIT_FIRMWARE))
split_firmware(master, part);
--- a/include/linux/mtd/partitions.h
+++ b/include/linux/mtd/partitions.h
@@ -74,6 +74,8 @@ struct mtd_part_parser_data {
enum mtd_parser_type {
MTD_PARSER_TYPE_DEVICE = 0,
+ MTD_PARSER_TYPE_ROOTFS,
+ MTD_PARSER_TYPE_FIRMWARE,
};
struct mtd_part_parser {

32
target/linux/generic/pending-4.14/403-mtd-hook-mtdsplit-to-Kbuild.patch Normal file
View file

@ -0,0 +1,32 @@
From: Gabor Juhos <juhosg@openwrt.org>
Subject: [PATCH] kernel/3.10: move squashfs check from rootfs split code into a separate file
lede-commit: d89bea92b31b4e157a0fa438e75370f089f73427
Signed-off-by: Gabor Juhos <juhosg@openwrt.org>
---
drivers/mtd/Kconfig | 2 ++
drivers/mtd/Makefile | 2 ++
2 files changed, 4 insertions(+)
--- a/drivers/mtd/Kconfig
+++ b/drivers/mtd/Kconfig
@@ -27,6 +27,8 @@ config MTD_SPLIT_FIRMWARE_NAME
depends on MTD_SPLIT_FIRMWARE
default "firmware"
+source "drivers/mtd/mtdsplit/Kconfig"
+
endmenu
config MTD_TESTS
--- a/drivers/mtd/Makefile
+++ b/drivers/mtd/Makefile
@@ -7,6 +7,8 @@
obj-$(CONFIG_MTD) += mtd.o
mtd-y := mtdcore.o mtdsuper.o mtdconcat.o mtdpart.o mtdchar.o
+obj-$(CONFIG_MTD_SPLIT) += mtdsplit/
+
obj-$(CONFIG_MTD_OF_PARTS) += ofpart.o
obj-$(CONFIG_MTD_REDBOOT_PARTS) += redboot.o
obj-$(CONFIG_MTD_CMDLINE_PARTS) += cmdlinepart.o

94
target/linux/generic/pending-4.14/404-mtd-add-more-helper-functions.patch Normal file
View file

@ -0,0 +1,94 @@
From: Gabor Juhos <juhosg@openwrt.org>
Subject: kernel/3.10: add separate rootfs partition parser
lede-commit: daec7ad7688415156e2730e401503d09bd3acf91
Signed-off-by: Gabor Juhos <juhosg@openwrt.org>
---
drivers/mtd/mtdpart.c | 29 +++++++++++++++++++++++++++++
include/linux/mtd/mtd.h | 18 ++++++++++++++++++
include/linux/mtd/partitions.h | 2 ++
3 files changed, 49 insertions(+)
--- a/drivers/mtd/mtdpart.c
+++ b/drivers/mtd/mtdpart.c
@@ -799,6 +799,17 @@ run_parsers_by_type(struct mtd_part *sla
return nr_parts;
}
+static inline unsigned long
+mtd_pad_erasesize(struct mtd_info *mtd, int offset, int len)
+{
+ unsigned long mask = mtd->erasesize - 1;
+
+ len += offset & mask;
+ len = (len + mask) & ~mask;
+ len -= offset & mask;
+ return len;
+}
+
#ifdef CONFIG_MTD_SPLIT_FIRMWARE_NAME
#define SPLIT_FIRMWARE_NAME CONFIG_MTD_SPLIT_FIRMWARE_NAME
#else
@@ -1144,6 +1155,24 @@ int mtd_is_partition(const struct mtd_in
}
EXPORT_SYMBOL_GPL(mtd_is_partition);
+struct mtd_info *mtdpart_get_master(const struct mtd_info *mtd)
+{
+ if (!mtd_is_partition(mtd))
+ return (struct mtd_info *)mtd;
+
+ return mtd_to_part(mtd)->parent;
+}
+EXPORT_SYMBOL_GPL(mtdpart_get_master);
+
+uint64_t mtdpart_get_offset(const struct mtd_info *mtd)
+{
+ if (!mtd_is_partition(mtd))
+ return 0;
+
+ return mtd_to_part(mtd)->offset;
+}
+EXPORT_SYMBOL_GPL(mtdpart_get_offset);
+
/* Returns the size of the entire flash chip */
uint64_t mtd_get_device_size(const struct mtd_info *mtd)
{
--- a/include/linux/mtd/mtd.h
+++ b/include/linux/mtd/mtd.h
@@ -493,6 +493,24 @@ static inline uint32_t mtd_mod_by_eb(uin
return do_div(sz, mtd->erasesize);
}
+static inline uint64_t mtd_roundup_to_eb(uint64_t sz, struct mtd_info *mtd)
+{
+ if (mtd_mod_by_eb(sz, mtd) == 0)
+ return sz;
+
+ /* Round up to next erase block */
+ return (mtd_div_by_eb(sz, mtd) + 1) * mtd->erasesize;
+}
+
+static inline uint64_t mtd_rounddown_to_eb(uint64_t sz, struct mtd_info *mtd)
+{
+ if (mtd_mod_by_eb(sz, mtd) == 0)
+ return sz;
+
+ /* Round down to the start of the current erase block */
+ return (mtd_div_by_eb(sz, mtd)) * mtd->erasesize;
+}
+
static inline uint32_t mtd_div_by_ws(uint64_t sz, struct mtd_info *mtd)
{
if (mtd->writesize_shift)
--- a/include/linux/mtd/partitions.h
+++ b/include/linux/mtd/partitions.h
@@ -114,6 +114,8 @@ int mtd_is_partition(const struct mtd_in
int mtd_add_partition(struct mtd_info *master, const char *name,
long long offset, long long length);
int mtd_del_partition(struct mtd_info *master, int partno);
+struct mtd_info *mtdpart_get_master(const struct mtd_info *mtd);
+uint64_t mtdpart_get_offset(const struct mtd_info *mtd);
uint64_t mtd_get_device_size(const struct mtd_info *mtd);
extern void __weak arch_split_mtd_part(struct mtd_info *master,
const char *name, int offset, int size);

154
target/linux/generic/pending-4.14/411-mtd-partial_eraseblock_write.patch Normal file
View file

@ -0,0 +1,154 @@
From: Felix Fietkau <nbd@nbd.name>
Subject: mtd: implement write support for partitions covering only a part of an eraseblock (buffer data that would otherwise be erased)
lede-commit: 87a8e8ac1067f58ba831c4aae443f3655c31cd80
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
drivers/mtd/mtdpart.c | 90 ++++++++++++++++++++++++++++++++++++++++++++-----
include/linux/mtd/mtd.h | 4 +++
2 files changed, 85 insertions(+), 9 deletions(-)
--- a/drivers/mtd/mtdpart.c
+++ b/drivers/mtd/mtdpart.c
@@ -36,6 +36,8 @@
#include "mtdcore.h"
#include "mtdsplit/mtdsplit.h"
+#define MTD_ERASE_PARTIAL 0x8000 /* partition only covers parts of an erase block */
+
/* Our partition linked list */
static LIST_HEAD(mtd_partitions);
static DEFINE_MUTEX(mtd_partitions_mutex);
@@ -241,13 +243,61 @@ static int part_erase(struct mtd_info *m
struct mtd_part *part = mtd_to_part(mtd);
int ret;
+
+ instr->partial_start = false;
+ if (mtd->flags & MTD_ERASE_PARTIAL) {
+ size_t readlen = 0;
+ u64 mtd_ofs;
+
+ instr->erase_buf = kmalloc(part->parent->erasesize, GFP_ATOMIC);
+ if (!instr->erase_buf)
+ return -ENOMEM;
+
+ mtd_ofs = part->offset + instr->addr;
+ instr->erase_buf_ofs = do_div(mtd_ofs, part->parent->erasesize);
+
+ if (instr->erase_buf_ofs > 0) {
+ instr->addr -= instr->erase_buf_ofs;
+ ret = mtd_read(part->parent,
+ instr->addr + part->offset,
+ part->parent->erasesize,
+ &readlen, instr->erase_buf);
+
+ instr->len += instr->erase_buf_ofs;
+ instr->partial_start = true;
+ } else {
+ mtd_ofs = part->offset + part->mtd.size;
+ instr->erase_buf_ofs = part->parent->erasesize -
+ do_div(mtd_ofs, part->parent->erasesize);
+
+ if (instr->erase_buf_ofs > 0) {
+ instr->len += instr->erase_buf_ofs;
+ ret = mtd_read(part->parent,
+ part->offset + instr->addr +
+ instr->len - part->parent->erasesize,
+ part->parent->erasesize, &readlen,
+ instr->erase_buf);
+ } else {
+ ret = 0;
+ }
+ }
+ if (ret < 0) {
+ kfree(instr->erase_buf);
+ return ret;
+ }
+
+ }
+
instr->addr += part->offset;
ret = part->parent->_erase(part->parent, instr);
if (ret) {
if (instr->fail_addr != MTD_FAIL_ADDR_UNKNOWN)
instr->fail_addr -= part->offset;
instr->addr -= part->offset;
+ if (mtd->flags & MTD_ERASE_PARTIAL)
+ kfree(instr->erase_buf);
}
+
return ret;
}
@@ -255,6 +305,25 @@ void mtd_erase_callback(struct erase_inf
{
if (instr->mtd->_erase == part_erase) {
struct mtd_part *part = mtd_to_part(instr->mtd);
+ size_t wrlen = 0;
+
+ if (instr->mtd->flags & MTD_ERASE_PARTIAL) {
+ if (instr->partial_start) {
+ part->parent->_write(part->parent,
+ instr->addr, instr->erase_buf_ofs,
+ &wrlen, instr->erase_buf);
+ instr->addr += instr->erase_buf_ofs;
+ } else {
+ instr->len -= instr->erase_buf_ofs;
+ part->parent->_write(part->parent,
+ instr->addr + instr->len,
+ instr->erase_buf_ofs, &wrlen,
+ instr->erase_buf +
+ part->parent->erasesize -
+ instr->erase_buf_ofs);
+ }
+ kfree(instr->erase_buf);
+ }
if (instr->fail_addr != MTD_FAIL_ADDR_UNKNOWN)
instr->fail_addr -= part->offset;
@@ -598,19 +667,22 @@ static struct mtd_part *allocate_partiti
remainder = do_div(tmp, wr_alignment);
if ((slave->mtd.flags & MTD_WRITEABLE) && remainder) {
/* Doesn't start on a boundary of major erase size */
- /* FIXME: Let it be writable if it is on a boundary of
- * _minor_ erase size though */
- slave->mtd.flags &= ~MTD_WRITEABLE;
- printk(KERN_WARNING"mtd: partition \"%s\" doesn't start on an erase/write block boundary -- force read-only\n",
- part->name);
+ slave->mtd.flags |= MTD_ERASE_PARTIAL;
+ if (((u32)slave->mtd.size) > parent->erasesize)
+ slave->mtd.flags &= ~MTD_WRITEABLE;
+ else
+ slave->mtd.erasesize = slave->mtd.size;
}
- tmp = slave->mtd.size;
+ tmp = slave->offset + slave->mtd.size;
remainder = do_div(tmp, wr_alignment);
if ((slave->mtd.flags & MTD_WRITEABLE) && remainder) {
- slave->mtd.flags &= ~MTD_WRITEABLE;
- printk(KERN_WARNING"mtd: partition \"%s\" doesn't end on an erase/write block -- force read-only\n",
- part->name);
+ slave->mtd.flags |= MTD_ERASE_PARTIAL;
+
+ if ((u32)slave->mtd.size > parent->erasesize)
+ slave->mtd.flags &= ~MTD_WRITEABLE;
+ else
+ slave->mtd.erasesize = slave->mtd.size;
}
mtd_set_ooblayout(&slave->mtd, &part_ooblayout_ops);
--- a/include/linux/mtd/mtd.h
+++ b/include/linux/mtd/mtd.h
@@ -56,6 +56,10 @@ struct erase_info {
u_long priv;
u_char state;
struct erase_info *next;
+
+ u8 *erase_buf;
+ u32 erase_buf_ofs;
+ bool partial_start;
};
struct mtd_erase_region_info {

40
target/linux/generic/pending-4.14/412-mtd-partial_eraseblock_unlock.patch Normal file
View file

@ -0,0 +1,40 @@
From: Tim Harvey <tharvey@gateworks.com>
Subject: mtd: allow partial block unlock
This allows sysupgrade for devices such as the Gateworks Avila/Cambria
product families based on the ixp4xx using the redboot bootloader with
combined FIS directory and RedBoot config partitions on larger FLASH
devices with larger eraseblocks.
This second iteration of this patch addresses previous issues:
- whitespace breakage fixed
- unlock in all scenarios
- simplification and fix logic bug
[john@phrozen.org: this should be moved to the ixp4xx folder]
Signed-off-by: Tim Harvey <tharvey@gateworks.com>
---
drivers/mtd/mtdpart.c | 11 ++++++++++-
1 file changed, 10 insertions(+), 1 deletion(-)
--- a/drivers/mtd/mtdpart.c
+++ b/drivers/mtd/mtdpart.c
@@ -343,7 +343,16 @@ static int part_lock(struct mtd_info *mt
static int part_unlock(struct mtd_info *mtd, loff_t ofs, uint64_t len)
{
struct mtd_part *part = mtd_to_part(mtd);
- return part->parent->_unlock(part->parent, ofs + part->offset, len);
+
+ ofs += part->offset;
+
+ if (mtd->flags & MTD_ERASE_PARTIAL) {
+ /* round up len to next erasesize and round down offset to prev block */
+ len = (mtd_div_by_eb(len, part->parent) + 1) * part->parent->erasesize;
+ ofs &= ~(part->parent->erasesize - 1);
+ }
+
+ return part->parent->_unlock(part->parent, ofs, len);
}
static int part_is_locked(struct mtd_info *mtd, loff_t ofs, uint64_t len)

41
target/linux/generic/pending-4.14/420-mtd-redboot_space.patch Normal file
View file

@ -0,0 +1,41 @@
From: Felix Fietkau <nbd@nbd.name>
Subject: add patch for including unpartitioned space in the rootfs partition for redboot devices (if applicable)
[john@phrozen.org: used by ixp and others]
lede-commit: 394918851f84e4d00fa16eb900e7700e95091f00
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
drivers/mtd/redboot.c | 19 +++++++++++++------
1 file changed, 13 insertions(+), 6 deletions(-)
--- a/drivers/mtd/redboot.c
+++ b/drivers/mtd/redboot.c
@@ -265,14 +265,21 @@ static int parse_redboot_partitions(stru
#endif
names += strlen(names)+1;
-#ifdef CONFIG_MTD_REDBOOT_PARTS_UNALLOCATED
if(fl->next && fl->img->flash_base + fl->img->size + master->erasesize <= fl->next->img->flash_base) {
- i++;
- parts[i].offset = parts[i-1].size + parts[i-1].offset;
- parts[i].size = fl->next->img->flash_base - parts[i].offset;
- parts[i].name = nullname;
- }
+ if (!strcmp(parts[i].name, "rootfs")) {
+ parts[i].size = fl->next->img->flash_base;
+ parts[i].size &= ~(master->erasesize - 1);
+ parts[i].size -= parts[i].offset;
+#ifdef CONFIG_MTD_REDBOOT_PARTS_UNALLOCATED
+ nrparts--;
+ } else {
+ i++;
+ parts[i].offset = parts[i-1].size + parts[i-1].offset;
+ parts[i].size = fl->next->img->flash_base - parts[i].offset;
+ parts[i].name = nullname;
#endif
+ }
+ }
tmp_fl = fl;
fl = fl->next;
kfree(tmp_fl);

47
target/linux/generic/pending-4.14/430-mtd-add-myloader-partition-parser.patch Normal file
View file

@ -0,0 +1,47 @@
From: Florian Fainelli <f.fainelli@gmail.com>
Subject: Add myloader partition table parser
[john@phozen.org: shoud be upstreamable]
lede-commit: d8bf22859b51faa09d22c056fe221a45d2f7a3b8
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
---
drivers/mtd/Kconfig | 16 ++++++++++++++++
drivers/mtd/Makefile | 1 +
2 files changed, 17 insertions(+)
--- a/drivers/mtd/Kconfig
+++ b/drivers/mtd/Kconfig
@@ -178,6 +178,22 @@ menu "Partition parsers"
source "drivers/mtd/parsers/Kconfig"
endmenu
+config MTD_MYLOADER_PARTS
+ tristate "MyLoader partition parsing"
+ depends on ADM5120 || ATH25 || ATH79
+ ---help---
+ MyLoader is a bootloader which allows the user to define partitions
+ in flash devices, by putting a table in the second erase block
+ on the device, similar to a partition table. This table gives the
+ offsets and lengths of the user defined partitions.
+
+ If you need code which can detect and parse these tables, and
+ register MTD 'partitions' corresponding to each image detected,
+ enable this option.
+
+ You will still need the parsing functions to be called by the driver
+ for your particular device. It won't happen automatically.
+
comment "User Modules And Translation Layers"
#
--- a/drivers/mtd/Makefile
+++ b/drivers/mtd/Makefile
@@ -16,6 +16,7 @@ obj-$(CONFIG_MTD_AFS_PARTS) += afs.o
obj-$(CONFIG_MTD_AR7_PARTS) += ar7part.o
obj-$(CONFIG_MTD_BCM63XX_PARTS) += bcm63xxpart.o
obj-$(CONFIG_MTD_BCM47XX_PARTS) += bcm47xxpart.o
+obj-$(CONFIG_MTD_MYLOADER_PARTS) += myloader.o
obj-y += parsers/
# 'Users' - code which presents functionality to userspace.

67
target/linux/generic/pending-4.14/431-mtd-bcm47xxpart-check-for-bad-blocks-when-calculatin.patch Normal file
View file

@ -0,0 +1,67 @@
From: =?UTF-8?q?Rafa=C5=82=20Mi=C5=82ecki?= <zajec5@gmail.com>
Subject: [PATCH] mtd: bcm47xxpart: check for bad blocks when calculating offsets
Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
---
--- a/drivers/mtd/parsers/parser_trx.c
+++ b/drivers/mtd/parsers/parser_trx.c
@@ -29,6 +29,33 @@ struct trx_header {
uint32_t offset[3];
} __packed;
+/*
+ * Calculate real end offset (address) for a given amount of data. It checks
+ * all blocks skipping bad ones.
+ */
+static size_t parser_trx_real_offset(struct mtd_info *mtd, size_t bytes)
+{
+ size_t real_offset = 0;
+
+ if (mtd_block_isbad(mtd, real_offset))
+ pr_warn("Base offset shouldn't be at bad block");
+
+ while (bytes >= mtd->erasesize) {
+ bytes -= mtd->erasesize;
+ real_offset += mtd->erasesize;
+ while (mtd_block_isbad(mtd, real_offset)) {
+ real_offset += mtd->erasesize;
+
+ if (real_offset >= mtd->size)
+ return real_offset - mtd->erasesize;
+ }
+ }
+
+ real_offset += bytes;
+
+ return real_offset;
+}
+
static const char *parser_trx_data_part_name(struct mtd_info *master,
size_t offset)
{
@@ -83,21 +110,21 @@ static int parser_trx_parse(struct mtd_i
if (trx.offset[2]) {
part = &parts[curr_part++];
part->name = "loader";
- part->offset = trx.offset[i];
+ part->offset = parser_trx_real_offset(mtd, trx.offset[i]);
i++;
}
if (trx.offset[i]) {
part = &parts[curr_part++];
part->name = "linux";
- part->offset = trx.offset[i];
+ part->offset = parser_trx_real_offset(mtd, trx.offset[i]);
i++;
}
if (trx.offset[i]) {
part = &parts[curr_part++];
part->name = parser_trx_data_part_name(mtd, trx.offset[i]);
- part->offset = trx.offset[i];
+ part->offset = parser_trx_real_offset(mtd, trx.offset[i]);
i++;
}

37
target/linux/generic/pending-4.14/432-mtd-bcm47xxpart-detect-T_Meter-partition.patch Normal file
View file

@ -0,0 +1,37 @@
From: =?UTF-8?q?Rafa=C5=82=20Mi=C5=82ecki?= <zajec5@gmail.com>
Subject: mtd: bcm47xxpart: detect T_Meter partition
It can be found on many Netgear devices. It consists of many 0x30 blocks
starting with 4D 54.
Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
---
drivers/mtd/bcm47xxpart.c | 10 ++++++++++
1 file changed, 10 insertions(+)
--- a/drivers/mtd/bcm47xxpart.c
+++ b/drivers/mtd/bcm47xxpart.c
@@ -39,6 +39,7 @@
#define NVRAM_HEADER 0x48534C46 /* FLSH */
#define POT_MAGIC1 0x54544f50 /* POTT */
#define POT_MAGIC2 0x504f /* OP */
+#define T_METER_MAGIC 0x4D540000 /* MT */
#define ML_MAGIC1 0x39685a42
#define ML_MAGIC2 0x26594131
#define TRX_MAGIC 0x30524448
@@ -182,6 +183,15 @@ static int bcm47xxpart_parse(struct mtd_
MTD_WRITEABLE);
continue;
}
+
+ /* T_Meter */
+ if ((le32_to_cpu(buf[0x000 / 4]) & 0xFFFF0000) == T_METER_MAGIC &&
+ (le32_to_cpu(buf[0x030 / 4]) & 0xFFFF0000) == T_METER_MAGIC &&
+ (le32_to_cpu(buf[0x060 / 4]) & 0xFFFF0000) == T_METER_MAGIC) {
+ bcm47xxpart_add_part(&parts[curr_part++], "T_Meter", offset,
+ MTD_WRITEABLE);
+ continue;
+ }
/* TRX */
if (buf[0x000 / 4] == TRX_MAGIC) {

116
target/linux/generic/pending-4.14/440-block2mtd_init.patch Normal file
View file

@ -0,0 +1,116 @@
From: Felix Fietkau <nbd@nbd.name>
Subject: block2mtd
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
drivers/mtd/devices/block2mtd.c | 30 ++++++++++++++++++++----------
1 file changed, 20 insertions(+), 10 deletions(-)
--- a/drivers/mtd/devices/block2mtd.c
+++ b/drivers/mtd/devices/block2mtd.c
@@ -26,6 +26,7 @@
#include <linux/list.h>
#include <linux/init.h>
#include <linux/mtd/mtd.h>
+#include <linux/mtd/partitions.h>
#include <linux/mutex.h>
#include <linux/mount.h>
#include <linux/slab.h>
@@ -219,7 +220,7 @@ static void block2mtd_free_device(struct
static struct block2mtd_dev *add_device(char *devname, int erase_size,
- int timeout)
+ const char *mtdname, int timeout)
{
#ifndef MODULE
int i;
@@ -227,6 +228,7 @@ static struct block2mtd_dev *add_device(
const fmode_t mode = FMODE_READ | FMODE_WRITE | FMODE_EXCL;
struct block_device *bdev = ERR_PTR(-ENODEV);
struct block2mtd_dev *dev;
+ struct mtd_partition *part;
char *name;
if (!devname)
@@ -283,13 +285,16 @@ static struct block2mtd_dev *add_device(
/* Setup the MTD structure */
/* make the name contain the block device in */
- name = kasprintf(GFP_KERNEL, "block2mtd: %s", devname);
+ if (!mtdname)
+ mtdname = devname;
+ name = kmalloc(strlen(mtdname) + 1, GFP_KERNEL);
if (!name)
goto err_destroy_mutex;
+ strcpy(name, mtdname);
dev->mtd.name = name;
- dev->mtd.size = dev->blkdev->bd_inode->i_size & PAGE_MASK;
+ dev->mtd.size = dev->blkdev->bd_inode->i_size & PAGE_MASK & ~(erase_size - 1);
dev->mtd.erasesize = erase_size;
dev->mtd.writesize = 1;
dev->mtd.writebufsize = PAGE_SIZE;
@@ -302,7 +307,11 @@ static struct block2mtd_dev *add_device(
dev->mtd.priv = dev;
dev->mtd.owner = THIS_MODULE;
- if (mtd_device_register(&dev->mtd, NULL, 0)) {
+ part = kzalloc(sizeof(struct mtd_partition), GFP_KERNEL);
+ part->name = name;
+ part->offset = 0;
+ part->size = dev->mtd.size;
+ if (mtd_device_register(&dev->mtd, part, 1)) {
/* Device didn't get added, so free the entry */
goto err_destroy_mutex;
}
@@ -310,8 +319,7 @@ static struct block2mtd_dev *add_device(
list_add(&dev->list, &blkmtd_device_list);
pr_info("mtd%d: [%s] erase_size = %dKiB [%d]\n",
dev->mtd.index,
- dev->mtd.name + strlen("block2mtd: "),
- dev->mtd.erasesize >> 10, dev->mtd.erasesize);
+ mtdname, dev->mtd.erasesize >> 10, dev->mtd.erasesize);
return dev;
err_destroy_mutex:
@@ -384,7 +392,7 @@ static int block2mtd_setup2(const char *
/* 80 for device, 12 for erase size, 80 for name, 8 for timeout */
char buf[80 + 12 + 80 + 8];
char *str = buf;
- char *token[2];
+ char *token[3];
char *name;
size_t erase_size = PAGE_SIZE;
unsigned long timeout = MTD_DEFAULT_TIMEOUT;
@@ -398,7 +406,7 @@ static int block2mtd_setup2(const char *
strcpy(str, val);
kill_final_newline(str);
- for (i = 0; i < 2; i++)
+ for (i = 0; i < 3; i++)
token[i] = strsep(&str, ",");
if (str) {
@@ -424,8 +432,10 @@ static int block2mtd_setup2(const char *
return 0;
}
}
+ if (token[2] && (strlen(token[2]) + 1 > 80))
+ pr_err("mtd device name too long\n");
- add_device(name, erase_size, timeout);
+ add_device(name, erase_size, token[2], timeout);
return 0;
}
@@ -459,7 +469,7 @@ static int block2mtd_setup(const char *v
module_param_call(block2mtd, block2mtd_setup, NULL, NULL, 0200);
-MODULE_PARM_DESC(block2mtd, "Device to use. \"block2mtd=<dev>[,<erasesize>]\"");
+MODULE_PARM_DESC(block2mtd, "Device to use. \"block2mtd=<dev>[,<erasesize>[,<name>]]\"");
static int __init block2mtd_init(void)
{

47
target/linux/generic/pending-4.14/441-block2mtd_probe.patch Normal file
View file

@ -0,0 +1,47 @@
From: Felix Fietkau <nbd@nbd.name>
Subject: block2mtd
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
drivers/mtd/devices/block2mtd.c | 9 ++++++---
1 file changed, 6 insertions(+), 3 deletions(-)
--- a/drivers/mtd/devices/block2mtd.c
+++ b/drivers/mtd/devices/block2mtd.c
@@ -392,7 +392,7 @@ static int block2mtd_setup2(const char *
/* 80 for device, 12 for erase size, 80 for name, 8 for timeout */
char buf[80 + 12 + 80 + 8];
char *str = buf;
- char *token[3];
+ char *token[4];
char *name;
size_t erase_size = PAGE_SIZE;
unsigned long timeout = MTD_DEFAULT_TIMEOUT;
@@ -406,7 +406,7 @@ static int block2mtd_setup2(const char *
strcpy(str, val);
kill_final_newline(str);
- for (i = 0; i < 3; i++)
+ for (i = 0; i < 4; i++)
token[i] = strsep(&str, ",");
if (str) {
@@ -435,6 +435,9 @@ static int block2mtd_setup2(const char *
if (token[2] && (strlen(token[2]) + 1 > 80))
pr_err("mtd device name too long\n");
+ if (token[3] && kstrtoul(token[3], 0, &timeout))
+ pr_err("invalid timeout\n");
+
add_device(name, erase_size, token[2], timeout);
return 0;
@@ -469,7 +472,7 @@ static int block2mtd_setup(const char *v
module_param_call(block2mtd, block2mtd_setup, NULL, NULL, 0200);
-MODULE_PARM_DESC(block2mtd, "Device to use. \"block2mtd=<dev>[,<erasesize>[,<name>]]\"");
+MODULE_PARM_DESC(block2mtd, "Device to use. \"block2mtd=<dev>[,<erasesize>[,<name>[,<timeout>]]]\"");
static int __init block2mtd_init(void)
{

25
target/linux/generic/pending-4.14/460-mtd-cfi_cmdset_0002-no-erase_suspend.patch Normal file
View file

@ -0,0 +1,25 @@
From: Felix Fietkau <nbd@nbd.name>
Subject: kernel: disable cfi cmdset 0002 erase suspend
on some platforms, erase suspend leads to data corruption and lockups when write
ops collide with erase ops. this has been observed on the buffalo wzr-hp-g300nh.
rather than play whack-a-mole with a hard to reproduce issue on a variety of devices,
simply disable erase suspend, as it will usually not produce any useful gain on
the small filesystems used on embedded hardware.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
drivers/mtd/chips/cfi_cmdset_0002.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
--- a/drivers/mtd/chips/cfi_cmdset_0002.c
+++ b/drivers/mtd/chips/cfi_cmdset_0002.c
@@ -811,7 +811,7 @@ static int get_chip(struct map_info *map
return 0;
case FL_ERASING:
- if (!cfip || !(cfip->EraseSuspend & (0x1|0x2)) ||
+ if (1 /* no suspend */ || !cfip || !(cfip->EraseSuspend & (0x1|0x2)) ||
!(mode == FL_READY || mode == FL_POINT ||
(mode == FL_WRITING && (cfip->EraseSuspend & 0x2))))
goto sleep;

17
target/linux/generic/pending-4.14/461-mtd-cfi_cmdset_0002-add-buffer-write-cmd-timeout.patch Normal file
View file

@ -0,0 +1,17 @@
From: George Kashperko <george@znau.edu.ua>
Subject: Issue map read after Write Buffer Load command to ensure chip is ready to receive data.
Signed-off-by: George Kashperko <george@znau.edu.ua>
---
drivers/mtd/chips/cfi_cmdset_0002.c | 1 +
1 file changed, 1 insertion(+)
--- a/drivers/mtd/chips/cfi_cmdset_0002.c
+++ b/drivers/mtd/chips/cfi_cmdset_0002.c
@@ -1832,6 +1832,7 @@ static int __xipram do_write_buffer(stru
/* Write Buffer Load */
map_write(map, CMD(0x25), cmd_adr);
+ (void) map_read(map, cmd_adr);
chip->state = FL_WRITING_TO_BUFFER;

18
target/linux/generic/pending-4.14/465-m25p80-mx-disable-software-protection.patch Normal file
View file

@ -0,0 +1,18 @@
From: Felix Fietkau <nbd@nbd.name>
Subject: Disable software protection bits for Macronix flashes.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
drivers/mtd/spi-nor/spi-nor.c | 1 +
1 file changed, 1 insertion(+)
--- a/drivers/mtd/spi-nor/spi-nor.c
+++ b/drivers/mtd/spi-nor/spi-nor.c
@@ -2715,6 +2715,7 @@ int spi_nor_scan(struct spi_nor *nor, co
if (JEDEC_MFR(info) == SNOR_MFR_ATMEL ||
JEDEC_MFR(info) == SNOR_MFR_INTEL ||
+ JEDEC_MFR(info) == SNOR_MFR_MACRONIX ||
JEDEC_MFR(info) == SNOR_MFR_SST ||
info->flags & SPI_NOR_HAS_LOCK) {
write_enable(nor);

56
target/linux/generic/pending-4.14/470-mtd-spi-nor-support-limiting-4K-sectors-support-base.patch Normal file
View file

@ -0,0 +1,56 @@
From: Felix Fietkau <nbd@nbd.name>
Date: Sat, 4 Nov 2017 07:40:23 +0100
Subject: [PATCH] mtd: spi-nor: support limiting 4K sectors support based on
flash size
Some devices need 4K sectors to be able to deal with small flash chips.
For instance, w25x05 is 64 KiB in size, and without 4K sectors, the
entire chip is just one erase block.
On bigger flash chip sizes, using 4K sectors can significantly slow down
many operations, including using a writable filesystem. There are several
platforms where it makes sense to use a single kernel on both kinds of
devices.
To support this properly, allow configuring an upper flash chip size
limit for 4K sectors support.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
--- a/drivers/mtd/spi-nor/Kconfig
+++ b/drivers/mtd/spi-nor/Kconfig
@@ -39,6 +39,17 @@ config SPI_ASPEED_SMC
and support for the SPI flash memory controller (SPI) for
the host firmware. The implementation only supports SPI NOR.
+config MTD_SPI_NOR_USE_4K_SECTORS_LIMIT
+ int "Maximum flash chip size to use 4K sectors on (in KiB)"
+ depends on MTD_SPI_NOR_USE_4K_SECTORS
+ default "4096"
+ help
+ There are many flash chips that support 4K sectors, but are so large
+ that using them significantly slows down writing large amounts of
+ data or using a writable filesystem.
+ Any flash chip larger than the size specified in this option will
+ not use 4K sectors.
+
config SPI_ATMEL_QUADSPI
tristate "Atmel Quad SPI Controller"
depends on ARCH_AT91 || (ARM && COMPILE_TEST)
--- a/drivers/mtd/spi-nor/spi-nor.c
+++ b/drivers/mtd/spi-nor/spi-nor.c
@@ -2562,10 +2562,12 @@ static int spi_nor_select_erase(struct s
#ifdef CONFIG_MTD_SPI_NOR_USE_4K_SECTORS
/* prefer "small sector" erase if possible */
- if (info->flags & SECT_4K) {
+ if ((info->flags & SECT_4K) && (mtd->size <=
+ CONFIG_MTD_SPI_NOR_USE_4K_SECTORS_LIMIT * 1024)) {
nor->erase_opcode = SPINOR_OP_BE_4K;
mtd->erasesize = 4096;
- } else if (info->flags & SECT_4K_PMC) {
+ } else if ((info->flags & SECT_4K_PMC) && (mtd->size <=
+ CONFIG_MTD_SPI_NOR_USE_4K_SECTORS_LIMIT * 1024)) {
nor->erase_opcode = SPINOR_OP_BE_4K_PMC;
mtd->erasesize = 4096;
} else

18
target/linux/generic/pending-4.14/476-mtd-spi-nor-add-eon-en25q128.patch Normal file
View file

@ -0,0 +1,18 @@
From: Piotr Dymacz <pepe2k@gmail.com>
Subject: kernel/mtd: add support for EON EN25Q128
Signed-off-by: Piotr Dymacz <pepe2k@gmail.com>
---
drivers/mtd/spi-nor/spi-nor.c | 1 +
1 file changed, 1 insertion(+)
--- a/drivers/mtd/spi-nor/spi-nor.c
+++ b/drivers/mtd/spi-nor/spi-nor.c
@@ -954,6 +954,7 @@ static const struct flash_info spi_nor_i
{ "en25q32b", INFO(0x1c3016, 0, 64 * 1024, 64, 0) },
{ "en25p64", INFO(0x1c2017, 0, 64 * 1024, 128, 0) },
{ "en25q64", INFO(0x1c3017, 0, 64 * 1024, 128, SECT_4K) },
+ { "en25q128", INFO(0x1c3018, 0, 64 * 1024, 256, SECT_4K) },
{ "en25qh128", INFO(0x1c7018, 0, 64 * 1024, 256, 0) },
{ "en25qh256", INFO(0x1c7019, 0, 64 * 1024, 512, 0) },
{ "en25s64", INFO(0x1c3817, 0, 64 * 1024, 128, SECT_4K) },

18
target/linux/generic/pending-4.14/477-mtd-add-spi-nor-add-mx25u3235f.patch Normal file
View file

@ -0,0 +1,18 @@
From: André Valentin <avalentin@marcant.net>
Subject: linux/mtd: add id for mx25u3235f needed by ZyXEL NBG6817
Signed-off-by: André Valentin <avalentin@marcant.net>
---
drivers/mtd/spi-nor/spi-nor.c | 1 +
1 file changed, 1 insertion(+)
--- a/drivers/mtd/spi-nor/spi-nor.c
+++ b/drivers/mtd/spi-nor/spi-nor.c
@@ -1016,6 +1016,7 @@ static const struct flash_info spi_nor_i
{ "mx25l3205d", INFO(0xc22016, 0, 64 * 1024, 64, SECT_4K) },
{ "mx25l3255e", INFO(0xc29e16, 0, 64 * 1024, 64, SECT_4K) },
{ "mx25l6405d", INFO(0xc22017, 0, 64 * 1024, 128, SECT_4K) },
+ { "mx25u3235f", INFO(0xc22536, 0, 64 * 1024, 64, 0) },
{ "mx25u2033e", INFO(0xc22532, 0, 64 * 1024, 4, SECT_4K) },
{ "mx25u4035", INFO(0xc22533, 0, 64 * 1024, 8, SECT_4K) },
{ "mx25u8035", INFO(0xc22534, 0, 64 * 1024, 16, SECT_4K) },

38
target/linux/generic/pending-4.14/480-mtd-set-rootfs-to-be-root-dev.patch Normal file
View file

@ -0,0 +1,38 @@
From: Gabor Juhos <juhosg@openwrt.org>
Subject: kernel/3.1[02]: move MTD root device setup code to mtdcore
The current code only allows to automatically set
root device on MTD partitions. Move the code to MTD
core to allow to use it with all MTD devices.
Signed-off-by: Gabor Juhos <juhosg@openwrt.org>
---
drivers/mtd/mtdcore.c | 10 ++++++++++
1 file changed, 10 insertions(+)
--- a/drivers/mtd/mtdcore.c
+++ b/drivers/mtd/mtdcore.c
@@ -41,6 +41,7 @@
#include <linux/reboot.h>
#include <linux/leds.h>
#include <linux/debugfs.h>
+#include <linux/root_dev.h>
#include <linux/mtd/mtd.h>
#include <linux/mtd/partitions.h>
@@ -578,6 +579,15 @@ int add_mtd_device(struct mtd_info *mtd)
of this try_ nonsense, and no bitching about it
either. :) */
__module_get(THIS_MODULE);
+
+ if (!strcmp(mtd->name, "rootfs") &&
+ IS_ENABLED(CONFIG_MTD_ROOTFS_ROOT_DEV) &&
+ ROOT_DEV == 0) {
+ pr_notice("mtd: device %d (%s) set to be root filesystem\n",
+ mtd->index, mtd->name);
+ ROOT_DEV = MKDEV(MTD_BLOCK_MAJOR, mtd->index);
+ }
+
return 0;
fail_added:

Some files were not shown because too many files have changed in this diff Show more