From: Willy Tarreau <w@1wt.eu>
To: linux-kernel@vger.kernel.org, stable@vger.kernel.org
Cc: Lars Persson <larper@axis.com>,
linux-mips@linux-mips.org, paul.burton@imgtec.com,
Ralf Baechle <ralf@linux-mips.org>,
Ben Hutchings <ben@decadent.org.uk>, Willy Tarreau <w@1wt.eu>
Subject: [PATCH 2.6.32 34/62] MIPS: Fix race condition in lazy cache flushing.
Date: Sun, 13 Sep 2015 00:56:40 +0200 [thread overview]
Message-ID: <20150912225608.028958489@1wt.eu> (raw)
In-Reply-To: <08d3b586eb2e764308c3de9ee398a17c@local>
2.6.32-longterm review patch. If anyone has any objections, please let me know.
------------------
From: Lars Persson <lars.persson@axis.com>
commit 4d46a67a3eb827ccf1125959936fd51ba318dabc upstream.
The lazy cache flushing implemented in the MIPS kernel suffers from a
race condition that is exposed by do_set_pte() in mm/memory.c.
A pre-condition is a file-system that writes to the page from the CPU
in its readpage method and then calls flush_dcache_page(). One example
is ubifs. Another pre-condition is that the dcache flush is postponed
in __flush_dcache_page().
Upon a page fault for an executable mapping not existing in the
page-cache, the following will happen:
1. Write to the page
2. flush_dcache_page
3. flush_icache_page
4. set_pte_at
5. update_mmu_cache (commits the flush of a dcache-dirty page)
Between steps 4 and 5 another thread can hit the same page and it will
encounter a valid pte. Because the data still is in the L1 dcache the CPU
will fetch stale data from L2 into the icache and execute garbage.
This fix moves the commit of the cache flush to step 3 to close the
race window. It also reduces the amount of flushes on non-executable
mappings because we never enter __flush_dcache_page() for non-aliasing
CPUs.
Regressions can occur in drivers that mistakenly relies on the
flush_dcache_page() in get_user_pages() for DMA operations.
[ralf@linux-mips.org: Folded in patch 9346 to fix highmem issue.]
Signed-off-by: Lars Persson <larper@axis.com>
Cc: linux-mips@linux-mips.org
Cc: paul.burton@imgtec.com
Cc: linux-kernel@vger.kernel.org
Patchwork: https://patchwork.linux-mips.org/patch/9346/
Patchwork: https://patchwork.linux-mips.org/patch/9738/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
(cherry picked from commit 6bde6a3df0b4c8680d51c987d446b0ff2d6df0a6)
Signed-off-by: Willy Tarreau <w@1wt.eu>
---
arch/mips/include/asm/cacheflush.h | 38 +++++++++++++++++++++++---------------
arch/mips/mm/cache.c | 12 ++++++++++++
2 files changed, 35 insertions(+), 15 deletions(-)
diff --git a/arch/mips/include/asm/cacheflush.h b/arch/mips/include/asm/cacheflush.h
index 03b1d69..2211f24 100644
--- a/arch/mips/include/asm/cacheflush.h
+++ b/arch/mips/include/asm/cacheflush.h
@@ -29,6 +29,20 @@
* - flush_icache_all() flush the entire instruction cache
* - flush_data_cache_page() flushes a page from the data cache
*/
+
+ /*
+ * This flag is used to indicate that the page pointed to by a pte
+ * is dirty and requires cleaning before returning it to the user.
+ */
+#define PG_dcache_dirty PG_arch_1
+
+#define Page_dcache_dirty(page) \
+ test_bit(PG_dcache_dirty, &(page)->flags)
+#define SetPageDcacheDirty(page) \
+ set_bit(PG_dcache_dirty, &(page)->flags)
+#define ClearPageDcacheDirty(page) \
+ clear_bit(PG_dcache_dirty, &(page)->flags)
+
extern void (*flush_cache_all)(void);
extern void (*__flush_cache_all)(void);
extern void (*flush_cache_mm)(struct mm_struct *mm);
@@ -37,12 +51,14 @@ extern void (*flush_cache_range)(struct vm_area_struct *vma,
unsigned long start, unsigned long end);
extern void (*flush_cache_page)(struct vm_area_struct *vma, unsigned long page, unsigned long pfn);
extern void __flush_dcache_page(struct page *page);
+extern void __flush_icache_page(struct vm_area_struct *vma, struct page *page);
static inline void flush_dcache_page(struct page *page)
{
- if (cpu_has_dc_aliases || !cpu_has_ic_fills_f_dc)
+ if (cpu_has_dc_aliases)
__flush_dcache_page(page);
-
+ else if (!cpu_has_ic_fills_f_dc)
+ SetPageDcacheDirty(page);
}
#define flush_dcache_mmap_lock(mapping) do { } while (0)
@@ -60,6 +76,11 @@ static inline void flush_anon_page(struct vm_area_struct *vma,
static inline void flush_icache_page(struct vm_area_struct *vma,
struct page *page)
{
+ if (!cpu_has_ic_fills_f_dc && (vma->vm_flags & VM_EXEC) &&
+ Page_dcache_dirty(page)) {
+ __flush_icache_page(vma, page);
+ ClearPageDcacheDirty(page);
+ }
}
extern void (*flush_icache_range)(unsigned long start, unsigned long end);
@@ -94,19 +115,6 @@ extern void (*flush_icache_all)(void);
extern void (*local_flush_data_cache_page)(void * addr);
extern void (*flush_data_cache_page)(unsigned long addr);
-/*
- * This flag is used to indicate that the page pointed to by a pte
- * is dirty and requires cleaning before returning it to the user.
- */
-#define PG_dcache_dirty PG_arch_1
-
-#define Page_dcache_dirty(page) \
- test_bit(PG_dcache_dirty, &(page)->flags)
-#define SetPageDcacheDirty(page) \
- set_bit(PG_dcache_dirty, &(page)->flags)
-#define ClearPageDcacheDirty(page) \
- clear_bit(PG_dcache_dirty, &(page)->flags)
-
/* Run kernel code uncached, useful for cache probing functions. */
unsigned long run_uncached(void *func);
diff --git a/arch/mips/mm/cache.c b/arch/mips/mm/cache.c
index 694d51f..37603a4 100644
--- a/arch/mips/mm/cache.c
+++ b/arch/mips/mm/cache.c
@@ -113,6 +113,18 @@ void __flush_anon_page(struct page *page, unsigned long vmaddr)
EXPORT_SYMBOL(__flush_anon_page);
+void __flush_icache_page(struct vm_area_struct *vma, struct page *page)
+{
+ unsigned long addr;
+
+ if (PageHighMem(page))
+ return;
+
+ addr = (unsigned long) page_address(page);
+ flush_data_cache_page(addr);
+}
+EXPORT_SYMBOL_GPL(__flush_icache_page);
+
void __update_cache(struct vm_area_struct *vma, unsigned long address,
pte_t pte)
{
--
1.7.12.2.21.g234cd45.dirty
next prev parent reply other threads:[~2015-09-12 23:12 UTC|newest]
Thread overview: 84+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <08d3b586eb2e764308c3de9ee398a17c@local>
2015-09-12 22:56 ` [PATCH 2.6.32 00/62] 2.6.32.68-longterm review Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 01/62] pipe: iovec: Fix memory corruption when retrying atomic copy as non-atomic Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 02/62] sg_start_req(): make sure that theres not too many elements in iovec Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 03/62] crypto: testmgr - update LZO compression test vectors Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 04/62] TTY: drop driver reference in tty_open fail path Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 05/62] netlink: fix possible spoofing from non-root processes Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 06/62] eCryptfs: Remove buggy and unnecessary write in file name decode routine Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 07/62] HID: fix a couple of off-by-ones Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 08/62] udf: Verify i_size when loading inode Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 09/62] udf: Verify symlink size before loading it Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 11/62] udf: Check path length when reading symlink Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 12/62] udf: Check component length before reading it Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 13/62] Remove repeated loads blocksize Willy Tarreau
2015-09-15 1:42 ` Ben Hutchings
2015-09-15 7:39 ` Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 14/62] udf: Check length of extended attributes and allocation descriptors Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 15/62] udp: fix behavior of wrong checksums Willy Tarreau
2015-09-15 1:44 ` Ben Hutchings
2015-09-15 7:41 ` Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 16/62] e1000: add dummy allocator to fix race condition between mtu change and netpoll Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 17/62] powerpc: Fix missing L2 cache size in /sys/devices/system/cpu Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 18/62] s390/hibernate: fix save and restore of kernel text section Willy Tarreau
2015-09-15 2:10 ` Ben Hutchings
2015-09-15 6:09 ` Heiko Carstens
2015-09-15 7:41 ` Willy Tarreau
2015-09-15 7:44 ` Heiko Carstens
2015-09-12 22:56 ` [PATCH 2.6.32 19/62] ptrace: fix race between ptrace_resume() and wait_task_stopped() Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 20/62] memstick: mspro_block: add missing curly braces Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 21/62] md/raid5: dont record new size if resize_stripes fails Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 22/62] powerpc: Align TOC to 256 bytes Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 23/62] jbd2: fix r_count overflows leading to buffer overflow in journal recovery Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 24/62] sd: Disable support for 256 byte/sector disks Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 25/62] lguest: fix out-by-one error in address checking Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 26/62] fs, omfs: add NULL terminator in the end up the token list Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 27/62] x86_64: Fix strnlen_user() to not touch memory after specified maximum Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 28/62] tracing: Have filter check for balanced ops Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 29/62] sctp: Fix race between OOTB responce and route removal Willy Tarreau
2015-09-15 2:26 ` Ben Hutchings
2015-09-15 7:43 ` Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 30/62] x86/reboot: Fix a warning message triggered by stop_other_cpus() Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 31/62] include/linux/sched.h: dont use task->pid/tgid in same_thread_group/has_group_leader_pid Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 32/62] __ptrace_may_access() should not deny sub-threads Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 33/62] MIPS: Octeon: Remove udelay() causing huge IRQ latency Willy Tarreau
2015-09-12 22:56 ` Willy Tarreau [this message]
2015-09-12 22:56 ` [PATCH 2.6.32 35/62] MIPS: Fix cpu_has_mips_r2_exec_hazard Willy Tarreau
2015-09-15 3:02 ` Ben Hutchings
2015-09-15 7:43 ` Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 36/62] MIPS: Octeon: Delete override of cpu_has_mips_r2_exec_hazard Willy Tarreau
2015-09-15 11:37 ` Ben Hutchings
2015-09-16 5:17 ` Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 37/62] hrtimer: Allow concurrent hrtimer_start() for self restarting timers Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 38/62] s5h1420: fix a buffer overflow when checking userspace params Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 39/62] cx24116: " Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 40/62] pktgen: adjust spacing in proc file interface output Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 41/62] staging: vt6655: device_rx_srv check sk_buff is NULL Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 42/62] fixing infinite OPEN loop in 4.0 stateid recovery Willy Tarreau
2015-09-14 23:54 ` Kornievskaia, Olga
2015-09-15 5:37 ` Willy Tarreau
[not found] ` <7E69BA76-E3DF-4389-A8C4-F23C9E1FD5CC@netapp.com>
[not found] ` <55D3ECC9-EC69-469F-AD46-EE3818F3D138@netapp.com>
2015-09-16 5:33 ` Willy Tarreau
2015-11-15 0:53 ` [stable] Failing to send a CLOSE if file is opened WRONLY and server reboots on a 4.x mount Ben Hutchings
2015-11-15 7:20 ` Willy Tarreau
2015-11-18 23:20 ` Luis Henriques
2015-09-12 22:56 ` [PATCH 2.6.32 43/62] SUNRPC: Fix a memory leak in the backchannel code Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 44/62] dmaengine: mv_xor: bug fix for racing condition in descriptors cleanup Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 45/62] ext4: fix race between truncate and __ext4_journalled_writepage() Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 46/62] Disable write buffering on Toshiba ToPIC95 Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 47/62] nfs: increase size of EXCHANGE_ID name string buffer Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 48/62] ext4: call sync_blockdev() before invalidate_bdev() in put_super() Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 49/62] NET: ROSE: Dont dereference NULL neighbour pointer Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 50/62] mm: kmemleak: allow safe memory scanning during kmemleak disabling Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 51/62] tracing/filter: Do not WARN on operand count going below zero Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 52/62] tracing/filter: Do not allow infix to exceed end of string Willy Tarreau
2015-09-12 22:56 ` [PATCH 2.6.32 53/62] fuse: initialize fc->release before calling it Willy Tarreau
2015-09-12 22:57 ` [PATCH 2.6.32 54/62] mm: avoid setting up anonymous pages into file mapping Willy Tarreau
2015-09-12 22:57 ` [PATCH 2.6.32 55/62] s390/process: fix sfpc inline assembly Willy Tarreau
2015-09-12 22:57 ` [PATCH 2.6.32 56/62] libata: increase the timeout when setting transfer mode Willy Tarreau
2015-09-12 22:57 ` [PATCH 2.6.32 57/62] x86/xen: Probe target addresses in set_aliased_prot() before the hypercall Willy Tarreau
2015-09-12 22:57 ` [PATCH 2.6.32 58/62] dccp: fix auto-loading of dccp(_probe) Willy Tarreau
2015-09-12 22:57 ` [PATCH 2.6.32 60/62] dccp: catch failed request_module call in dccp_probe init Willy Tarreau
2015-09-12 22:57 ` [PATCH 2.6.32 61/62] dmaengine: fix missing cnt in ?: in dmatest Willy Tarreau
2015-09-12 22:57 ` [PATCH 2.6.32 62/62] ipv6: Fix return of xfrm6_tunnel_rcv() Willy Tarreau
2015-09-12 23:18 ` [PATCH 2.6.32 00/62] 2.6.32.68-longterm review Willy Tarreau
2015-09-15 12:06 ` Ben Hutchings
2015-09-16 5:23 ` Willy Tarreau
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20150912225608.028958489@1wt.eu \
--to=w@1wt.eu \
--cc=ben@decadent.org.uk \
--cc=larper@axis.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mips@linux-mips.org \
--cc=paul.burton@imgtec.com \
--cc=ralf@linux-mips.org \
--cc=stable@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).