Crash when use of the sd card + network card #52

alphaonex86 · 2012-07-11T22:31:08Z

Hello, I rsync the folder on the sd card (normal folder /usr/portage/), it use the intensive the network and sd card. After 10min of work all it crashed, I need hard reboot. I'm under console, and the linux put the back screen after inactivity, I can't press key board my keyboard not work (see the previous reported bug).
But I have noted lot of similar problem of stability when you use the sd card (read + write). My sd card is kingstone class 4 of 16GB.
Freeze mean for me: no responds to the ping.

popcornmix · 2012-07-11T22:42:16Z

What does "uname -a" report?

alphaonex86 · 2012-07-11T23:00:33Z

Linux toc 3.1.9 #3 PREEMPT Tue Jul 17 20:20:00 CEST 2012 armv6l ARMv6-compatible processor rev 7 (v6l) BCM2708 GNU/Linux
Then the official kernel, but with custom config (add some modules, remove other, but nothing important)

popcornmix · 2012-07-11T23:48:26Z

Can you confirm it happens with official build of kernel?
Can you capture a stack trace?

alphaonex86 · 2012-07-12T07:15:50Z

How capture the stack trace?

EDIT: On console, the cursor stop to blink

alphaonex86 · 2012-07-12T09:50:29Z

Linked only with the usb/network/general because I have crash too without disk access (I think: find ./ -mmin -60 say nothing, and lot of stuff is under tmpfs).

popcornmix · 2012-07-12T12:12:35Z

Normally stack trace is output to screen when a crash occurs.
Are you overclocking? Are you sure power supply is good?

alphaonex86 · 2012-07-12T12:25:29Z

Then like my normal computer.
No overclocking (official clock). I presume my power supply as good (provided by the raspberry pi resseller), but I have any tools to check it. I have nothing to weld an capacitor to test (and If I failed the weld, it's 30$ more + 450$ to ship where I live).
All my router/computer/device seam work here.

alphaonex86 · 2012-07-12T12:28:24Z

But I note red light up when the power is disconnected and hub powered connected (no diode to prevent alimenting by usb I presume).

alphaonex86 · 2012-07-12T12:32:02Z

I have a dmesg on my custom kernel:
Backtrace:
http://pastebin.com/01gtGESk

I search enable my symboles to debug better...

alphaonex86 · 2012-07-12T14:34:18Z

No bt but when I disconnect the network card, the rpi stop freeze and display in loop the message for the network card:
event key 5 dropped... I will note it.

alphaonex86 · 2012-07-12T14:50:57Z

To do a bug, it's very easy:
I launch 10x ping command like that's, on my personal computer on the lan:
ping www.google.com
my rpi is router/gateway, eth0 (internal interface) is for the lan 192.168.158.0/24, the eth1 is for the adsl routeur 192.168.1.0/24 (linked with gateway/masquerade mode?)

I have in loop on the hdmi output:
smsc95xx 1-1.1:1.0: eth0: kevent 4 may have been dropped
and unable to acces to the dmesg (I have try loop into screen session, dmesg > /boot/dmesg;sleep 60, with /boot as sync)

alphaonex86 · 2012-07-14T14:19:29Z

No crash the same device but with routing/gateway disabled.

confirmed by the bt::
http://pastebin.com/01gtGESk

alphaonex86 · 2012-07-16T09:18:36Z

Power is correct, I have changed all the device to be sure.
The back trace is clear, no? The card asix receive RX, then the data is transmited into an interruption to smsc95xx, and sometime that's do a infinity loop.

I need very quickly a solution, if you can do that's, thanks to you.

alphaonex86 · 2012-07-16T19:56:19Z

With CONFIG_CPU_IDLE that's do lot of time before the problem with my custom kernel, but the ping and bandwith to internet is more and more worse (I need reset each 6h). But at final it crash too.
Mem leak?

alphaonex86 · 2012-07-25T21:28:17Z

I'm without swap.

alphaonex86 · 2012-08-20T09:28:22Z

Now I have error:
Error::handle_he_chhltd_intr_dma:1988 handle_he_chhltd_intr_dma: Channel 3, DMA Mode -- but reason for halting is unknow, hcint 0x00000402, insts 0x0660001
Error::handle_he_chhltd_intr_dma:1988 handle_he_chhltd_intr_dma: Channel 3, DMA Mode -- but reason for halting is unknow, hcint 0x00000402, insts 0x0660009
Error::handle_he_chhltd_intr_dma:1988 handle_he_chhltd_intr_dma: Channel 3, DMA Mode -- but reason for halting is unknow, hcint 0x00000402, insts 0x0660021
in loop

popcornmix · 2012-08-20T21:18:48Z

Can you test with latest kernel. Also can you try with:
dwc_otg.microframe_schedule=1
in cmdline.txt

alphaonex86 · 2012-08-26T11:00:44Z

The bug is same, I have tryied with the last kernel at this time.

ashleyw · 2012-08-27T21:12:58Z

I had this exact problem @alphaonex86, "Channel 3, DMA Mode -- but reason for halting is unknown", caused by simply wget'ing a large file (it'd freeze after ~2 minutes.)

It turned out it was my Apple USB wired keyboard causing it — once I removed it, everything ran smoothly! I guess it was sucking too much power?

Others seem to agree: Report: Apple Wired Keyboard Incompatible

alphaonex86 · 2012-08-27T22:10:25Z

I have external power, then I my case it's not linked with power problem. But it seam like have infinity loop after disconnect the device (like when connected) and consume all the cpu.

For my it's context problem and software only bug.

alphaonex86 · 2012-10-10T13:22:45Z

Now I have:

[ 1201.006730] VCHIQka-0 D c0372168 0 23 2 0x00000000
[ 1201.013922] from
[ 1201.020866] from
[ 1201.027978] from
[ 1321.032970] INFO: task VCHIQka-0:23 blocked for more than 120 seconds.
[ 1321.039772] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 1321.062936] VCHIQka-0 D c0372168 0 23 2 0x00000000
[ 1321.069966] from
[ 1321.092995] from
[ 1321.100052] from

popcornmix · 2012-10-10T13:58:18Z

@alphaonex86
That is a known problem. See here:
http://www.raspberrypi.org/phpBB3/viewtopic.php?f=29&t=19334

alphaonex86 · 2012-10-10T21:41:12Z

I don't know if is linked with my problem, but before I was slow down and after freeze, now I have just slow down (but after more time, freeze after more time?)
The slow down is show by ping to rpi greater than 1000ms (the normal at starting is <1ms).
I can disable the videocore module if needed, I use kms (just to have console output) but no X.

alphaonex86 · 2012-10-11T12:48:00Z

Same, do just more time to bug (1 day).

alphaonex86 · 2012-10-19T14:34:15Z

I have try disable the turbo mode, with more large network buffer (problem after more time, ...), I have try large the memory kernel min size...

alphaonex86 · 2012-11-06T15:16:47Z

Same with the last kernel

ghollingworth · 2012-11-07T17:57:08Z

Which kernel revision?

"uname -a"

alphaonex86 · 2012-11-07T22:20:43Z

Not usefull information, because you don't have the real version:
Linux toc 3.2.27 #1 Mon Nov 5 03:14:21 CET 2012 armv6l ARMv6-compatible processor rev 7 (v6l) BCM2708 GNU/Linux
I have try at this commit:
df41b42

alphaonex86 · 2012-11-10T17:58:28Z

Better after this commit, due the previous commit on network card:
1aa222e
But remain crashed after few hour (about 8h with network trafic as gateway).

ghollingworth · 2012-11-10T18:46:51Z

Make sure turbo mode is N

On 10/11/2012, BRULE Herman notifications@github.com wrote:

Better with:
1aa222e
But remain crashed after few hour (about 8h with network trafic as
gateway).

Reply to this email directly or view it on GitHub:
#52 (comment)

Sent from my mobile device

alphaonex86 · 2012-11-10T19:05:03Z

I'm sure to that's. But I can check by sysctl or dmesg?

ghollingworth · 2013-07-17T20:31:05Z

Can you re-check with latest 3.6.y kernel, reopen if still found

MIPI Block #52 which provides configuration details for the MIPI panel including dphy settings as per panel and tcon specs Block #53 gives information on panel enable sequences v2: Address review comemnts from Jani - Move panel ids from intel_dsi.h to intel_bios.h - bdb_mipi_config structure improvements for cleaner code - Adding units for the pps delays, all in ms - change data structure to be more cleaner and simple v3: Corrected the unit for pps delays as 100us Signed-off-by: Shobhit Kumar <shobhit.kumar@intel.com> Reviewed-by: Jani Nikula <jani.nikula@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

The parser extracts the config block(#52) and sequence(#53) data and store in private data structures. v2: Address review comments by Jani - adjust code for the structure changes for bdb_mipi_config - add boundry and buffer overflow checks as suggested - use kmemdup instead of kmalloc and memcpy v3: More strict check while parsing VBT - Ensure that at anytime we do not go beyond sequence block while parsing - On unknown element fail the whole parsing v4: Style changes and spell check mostly as suggested by Jani Signed-off-by: Shobhit Kumar <shobhit.kumar@intel.com> Reviewed-by: Jani Nikula <jani.nikula@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

This driver makes use of the generic panel information from the VBT. Panel information is classified into two - panel configuration and panel power sequence which is unique to each panel. The generic driver uses the panel configuration and sequence parsed from VBT block #52 and #53 v2: Address review comments by Jani - Move all of the things in driver c file from header - Make all functions static - Make use of video/mipi_display.c instead of redefining - Null checks during sequence execution v3: Address review comments by Damien - Rename the panel driver file as intel_dsi_panel_vbt.c - Fix style changes as suggested - Correct comments for lp->hs and hs->lp count calculations - General updating comments to have more clarity - using max() instead of ternary operator - Fix names (ui_num, ui_den) while using UI in calculations - compute max of lp_to_hs switch and hs_to_lp switch while computing hs_lp_switch_count Signed-off-by: Shobhit Kumar <shobhit.kumar@intel.com> Reviewed-by: Damien Lespiau <damien.lespiau@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

…ode() In __ftrace_replace_code(), when converting the call to a nop in a function it needs to compare against the "curr" (current) value of the ftrace ops, and not the "new" one. It currently does not affect x86 which is the only arch to do the trampolines with function graph tracer, but when other archs that do depend on this code implement the function graph trampoline, it can crash. Here's an example when ARM uses the trampolines (in the future): ------------[ cut here ]------------ WARNING: CPU: 0 PID: 9 at kernel/trace/ftrace.c:1716 ftrace_bug+0x17c/0x1f4() Modules linked in: omap_rng rng_core ipv6 CPU: 0 PID: 9 Comm: migration/0 Not tainted 3.16.0-test-10959-gf0094b28f303-dirty #52 [<c02188f4>] (unwind_backtrace) from [<c021343c>] (show_stack+0x20/0x24) [<c021343c>] (show_stack) from [<c095a674>] (dump_stack+0x78/0x94) [<c095a674>] (dump_stack) from [<c02532a0>] (warn_slowpath_common+0x7c/0x9c) [<c02532a0>] (warn_slowpath_common) from [<c02532ec>] (warn_slowpath_null+0x2c/0x34) [<c02532ec>] (warn_slowpath_null) from [<c02cbac4>] (ftrace_bug+0x17c/0x1f4) [<c02cbac4>] (ftrace_bug) from [<c02cc44c>] (ftrace_replace_code+0x80/0x9c) [<c02cc44c>] (ftrace_replace_code) from [<c02cc658>] (ftrace_modify_all_code+0xb8/0x164) [<c02cc658>] (ftrace_modify_all_code) from [<c02cc718>] (__ftrace_modify_code+0x14/0x1c) [<c02cc718>] (__ftrace_modify_code) from [<c02c7244>] (multi_cpu_stop+0xf4/0x134) [<c02c7244>] (multi_cpu_stop) from [<c02c6e90>] (cpu_stopper_thread+0x54/0x130) [<c02c6e90>] (cpu_stopper_thread) from [<c0271cd4>] (smpboot_thread_fn+0x1ac/0x1bc) [<c0271cd4>] (smpboot_thread_fn) from [<c026ddf0>] (kthread+0xe0/0xfc) [<c026ddf0>] (kthread) from [<c020f318>] (ret_from_fork+0x14/0x20) ---[ end trace dc9ce72c5b617d8f ]--- [ 65.047264] ftrace failed to modify [<c0208580>] asm_do_IRQ+0x10/0x1c [ 65.054070] actual: 85:1b:00:eb Fixes: 7413af1 "ftrace: Make get_ftrace_addr() and get_ftrace_addr_old() global" Signed-off-by: Steven Rostedt <rostedt@goodmis.org>

…ode() commit 39b5552 upstream. In __ftrace_replace_code(), when converting the call to a nop in a function it needs to compare against the "curr" (current) value of the ftrace ops, and not the "new" one. It currently does not affect x86 which is the only arch to do the trampolines with function graph tracer, but when other archs that do depend on this code implement the function graph trampoline, it can crash. Here's an example when ARM uses the trampolines (in the future): ------------[ cut here ]------------ WARNING: CPU: 0 PID: 9 at kernel/trace/ftrace.c:1716 ftrace_bug+0x17c/0x1f4() Modules linked in: omap_rng rng_core ipv6 CPU: 0 PID: 9 Comm: migration/0 Not tainted 3.16.0-test-10959-gf0094b28f303-dirty #52 [<c02188f4>] (unwind_backtrace) from [<c021343c>] (show_stack+0x20/0x24) [<c021343c>] (show_stack) from [<c095a674>] (dump_stack+0x78/0x94) [<c095a674>] (dump_stack) from [<c02532a0>] (warn_slowpath_common+0x7c/0x9c) [<c02532a0>] (warn_slowpath_common) from [<c02532ec>] (warn_slowpath_null+0x2c/0x34) [<c02532ec>] (warn_slowpath_null) from [<c02cbac4>] (ftrace_bug+0x17c/0x1f4) [<c02cbac4>] (ftrace_bug) from [<c02cc44c>] (ftrace_replace_code+0x80/0x9c) [<c02cc44c>] (ftrace_replace_code) from [<c02cc658>] (ftrace_modify_all_code+0xb8/0x164) [<c02cc658>] (ftrace_modify_all_code) from [<c02cc718>] (__ftrace_modify_code+0x14/0x1c) [<c02cc718>] (__ftrace_modify_code) from [<c02c7244>] (multi_cpu_stop+0xf4/0x134) [<c02c7244>] (multi_cpu_stop) from [<c02c6e90>] (cpu_stopper_thread+0x54/0x130) [<c02c6e90>] (cpu_stopper_thread) from [<c0271cd4>] (smpboot_thread_fn+0x1ac/0x1bc) [<c0271cd4>] (smpboot_thread_fn) from [<c026ddf0>] (kthread+0xe0/0xfc) [<c026ddf0>] (kthread) from [<c020f318>] (ret_from_fork+0x14/0x20) ---[ end trace dc9ce72c5b617d8f ]--- [ 65.047264] ftrace failed to modify [<c0208580>] asm_do_IRQ+0x10/0x1c [ 65.054070] actual: 85:1b:00:eb Fixes: 7413af1 "ftrace: Make get_ftrace_addr() and get_ftrace_addr_old() global" Signed-off-by: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mem_cgroup_migrate() uses local_irq_disable/enable() but can be called with irq disabled from migrate_page_copy(). This ends up enabling irq while holding a irq context lock triggering the following lockdep warning. Fix it by using irq_save/restore instead. ================================= [ INFO: inconsistent lock state ] 4.7.0-rc1+ #52 Tainted: G W --------------------------------- inconsistent {IN-SOFTIRQ-W} -> {SOFTIRQ-ON-W} usage. kcompactd0/151 [HC0[0]:SC0[0]:HE1:SE1] takes: (&(&ctx->completion_lock)->rlock){+.?.-.}, at: [<000000000038fd96>] aio_migratepage+0x156/0x1e8 {IN-SOFTIRQ-W} state was registered at: __lock_acquire+0x5b6/0x1930 lock_acquire+0xee/0x270 _raw_spin_lock_irqsave+0x66/0xb0 aio_complete+0x98/0x328 dio_complete+0xe4/0x1e0 blk_update_request+0xd4/0x450 scsi_end_request+0x48/0x1c8 scsi_io_completion+0x272/0x698 blk_done_softirq+0xca/0xe8 __do_softirq+0xc8/0x518 irq_exit+0xee/0x110 do_IRQ+0x6a/0x88 io_int_handler+0x11a/0x25c __mutex_unlock_slowpath+0x144/0x1d8 __mutex_unlock_slowpath+0x140/0x1d8 kernfs_iop_permission+0x64/0x80 __inode_permission+0x9e/0xf0 link_path_walk+0x6e/0x510 path_lookupat+0xc4/0x1a8 filename_lookup+0x9c/0x160 user_path_at_empty+0x5c/0x70 SyS_readlinkat+0x68/0x140 system_call+0xd6/0x270 irq event stamp: 971410 hardirqs last enabled at (971409): migrate_page_move_mapping+0x3ea/0x588 hardirqs last disabled at (971410): _raw_spin_lock_irqsave+0x3c/0xb0 softirqs last enabled at (970526): __do_softirq+0x460/0x518 softirqs last disabled at (970519): irq_exit+0xee/0x110 other info that might help us debug this: Possible unsafe locking scenario: CPU0 ---- lock(&(&ctx->completion_lock)->rlock); <Interrupt> lock(&(&ctx->completion_lock)->rlock); *** DEADLOCK *** 3 locks held by kcompactd0/151: #0: (&(&mapping->private_lock)->rlock){+.+.-.}, at: aio_migratepage+0x42/0x1e8 #1: (&ctx->ring_lock){+.+.+.}, at: aio_migratepage+0x5a/0x1e8 #2: (&(&ctx->completion_lock)->rlock){+.?.-.}, at: aio_migratepage+0x156/0x1e8 stack backtrace: CPU: 20 PID: 151 Comm: kcompactd0 Tainted: G W 4.7.0-rc1+ #52 Call Trace: show_trace+0xea/0xf0 show_stack+0x72/0xf0 dump_stack+0x9a/0xd8 print_usage_bug.part.27+0x2d4/0x2e8 mark_lock+0x17e/0x758 mark_held_locks+0xa2/0xd0 trace_hardirqs_on_caller+0x140/0x1c0 mem_cgroup_migrate+0x266/0x370 aio_migratepage+0x16a/0x1e8 move_to_new_page+0xb0/0x260 migrate_pages+0x8f4/0x9f0 compact_zone+0x4dc/0xdc8 kcompactd_do_work+0x1aa/0x358 kcompactd+0xba/0x2c8 kthread+0x10a/0x110 kernel_thread_starter+0x6/0xc kernel_thread_starter+0x0/0xc INFO: lockdep is turned off. Link: http://lkml.kernel.org/r/20160620184158.GO3262@mtj.duckdns.org Link: http://lkml.kernel.org/g/5767CFE5.7080904@de.ibm.com Fixes: 74485cf ("mm: migrate: consolidate mem_cgroup_migrate() calls") Signed-off-by: Tejun Heo <tj@kernel.org> Reported-by: Christian Borntraeger <borntraeger@de.ibm.com> Acked-by: Johannes Weiner <hannes@cmpxchg.org> Acked-by: Michal Hocko <mhocko@suse.com> Reviewed-by: Vladimir Davydov <vdavydov@virtuozzo.com> Cc: <stable@vger.kernel.org> [4.5+] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

commit d93c413 upstream. mem_cgroup_migrate() uses local_irq_disable/enable() but can be called with irq disabled from migrate_page_copy(). This ends up enabling irq while holding a irq context lock triggering the following lockdep warning. Fix it by using irq_save/restore instead. ================================= [ INFO: inconsistent lock state ] 4.7.0-rc1+ #52 Tainted: G W --------------------------------- inconsistent {IN-SOFTIRQ-W} -> {SOFTIRQ-ON-W} usage. kcompactd0/151 [HC0[0]:SC0[0]:HE1:SE1] takes: (&(&ctx->completion_lock)->rlock){+.?.-.}, at: [<000000000038fd96>] aio_migratepage+0x156/0x1e8 {IN-SOFTIRQ-W} state was registered at: __lock_acquire+0x5b6/0x1930 lock_acquire+0xee/0x270 _raw_spin_lock_irqsave+0x66/0xb0 aio_complete+0x98/0x328 dio_complete+0xe4/0x1e0 blk_update_request+0xd4/0x450 scsi_end_request+0x48/0x1c8 scsi_io_completion+0x272/0x698 blk_done_softirq+0xca/0xe8 __do_softirq+0xc8/0x518 irq_exit+0xee/0x110 do_IRQ+0x6a/0x88 io_int_handler+0x11a/0x25c __mutex_unlock_slowpath+0x144/0x1d8 __mutex_unlock_slowpath+0x140/0x1d8 kernfs_iop_permission+0x64/0x80 __inode_permission+0x9e/0xf0 link_path_walk+0x6e/0x510 path_lookupat+0xc4/0x1a8 filename_lookup+0x9c/0x160 user_path_at_empty+0x5c/0x70 SyS_readlinkat+0x68/0x140 system_call+0xd6/0x270 irq event stamp: 971410 hardirqs last enabled at (971409): migrate_page_move_mapping+0x3ea/0x588 hardirqs last disabled at (971410): _raw_spin_lock_irqsave+0x3c/0xb0 softirqs last enabled at (970526): __do_softirq+0x460/0x518 softirqs last disabled at (970519): irq_exit+0xee/0x110 other info that might help us debug this: Possible unsafe locking scenario: CPU0 ---- lock(&(&ctx->completion_lock)->rlock); <Interrupt> lock(&(&ctx->completion_lock)->rlock); *** DEADLOCK *** 3 locks held by kcompactd0/151: #0: (&(&mapping->private_lock)->rlock){+.+.-.}, at: aio_migratepage+0x42/0x1e8 #1: (&ctx->ring_lock){+.+.+.}, at: aio_migratepage+0x5a/0x1e8 #2: (&(&ctx->completion_lock)->rlock){+.?.-.}, at: aio_migratepage+0x156/0x1e8 stack backtrace: CPU: 20 PID: 151 Comm: kcompactd0 Tainted: G W 4.7.0-rc1+ #52 Call Trace: show_trace+0xea/0xf0 show_stack+0x72/0xf0 dump_stack+0x9a/0xd8 print_usage_bug.part.27+0x2d4/0x2e8 mark_lock+0x17e/0x758 mark_held_locks+0xa2/0xd0 trace_hardirqs_on_caller+0x140/0x1c0 mem_cgroup_migrate+0x266/0x370 aio_migratepage+0x16a/0x1e8 move_to_new_page+0xb0/0x260 migrate_pages+0x8f4/0x9f0 compact_zone+0x4dc/0xdc8 kcompactd_do_work+0x1aa/0x358 kcompactd+0xba/0x2c8 kthread+0x10a/0x110 kernel_thread_starter+0x6/0xc kernel_thread_starter+0x0/0xc INFO: lockdep is turned off. Link: http://lkml.kernel.org/r/20160620184158.GO3262@mtj.duckdns.org Link: http://lkml.kernel.org/g/5767CFE5.7080904@de.ibm.com Fixes: 74485cf ("mm: migrate: consolidate mem_cgroup_migrate() calls") Signed-off-by: Tejun Heo <tj@kernel.org> Reported-by: Christian Borntraeger <borntraeger@de.ibm.com> Acked-by: Johannes Weiner <hannes@cmpxchg.org> Acked-by: Michal Hocko <mhocko@suse.com> Reviewed-by: Vladimir Davydov <vdavydov@virtuozzo.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

When interrupting an application which was allocating DMAable memory, it was possible, that the DMA memory was deallocated twice, leading to the error symptoms below. Thanks to Gerald, who analyzed the problem and provided this patch. I agree with his analysis of the problem: ddcb_cmd_fixups() -> genwqe_alloc_sync_sgl() (fails in f/lpage, but sgl->sgl != NULL and f/lpage maybe also != NULL) -> ddcb_cmd_cleanup() -> genwqe_free_sync_sgl() (double free, because sgl->sgl != NULL and f/lpage maybe also != NULL) In this scenario we would have exactly the kind of double free that would explain the WARNING / Bad page state, and as expected it is caused by broken error handling (cleanup). Using the Ubuntu git source, tag Ubuntu-4.4.0-33.52, he was able to reproduce the "Bad page state" issue, and with the patch on top he could not reproduce it any more. ------------[ cut here ]------------ WARNING: at /build/linux-o03cxz/linux-4.4.0/arch/s390/include/asm/pci_dma.h:141 Modules linked in: qeth_l2 ghash_s390 prng aes_s390 des_s390 des_generic sha512_s390 sha256_s390 sha1_s390 sha_common genwqe_card qeth crc_itu_t qdio ccwgroup vmur dm_multipath dasd_eckd_mod dasd_mod CPU: 2 PID: 3293 Comm: genwqe_gunzip Not tainted 4.4.0-33-generic #52-Ubuntu task: 0000000032c7e270 ti: 00000000324e4000 task.ti: 00000000324e4000 Krnl PSW : 0404c00180000000 0000000000156346 (dma_update_cpu_trans+0x9e/0xa8) R:0 T:1 IO:0 EX:0 Key:0 M:1 W:0 P:0 AS:3 CC:0 PM:0 EA:3 Krnl GPRS: 00000000324e7bcd 0000000000c3c34a 0000000027628298 000000003215b400 0000000000000400 0000000000001fff 0000000000000400 0000000116853000 07000000324e7b1e 0000000000000001 0000000000000001 0000000000000001 0000000000001000 0000000116854000 0000000000156402 00000000324e7a38 Krnl Code: 000000000015633a: 95001000 cli 0(%r1),0 000000000015633e: a774ffc3 brc 7,1562c4 #0000000000156342: a7f40001 brc 15,156344 >0000000000156346: 92011000 mvi 0(%r1),1 000000000015634a: a7f4ffbd brc 15,1562c4 000000000015634e: 0707 bcr 0,%r7 0000000000156350: c00400000000 brcl 0,156350 0000000000156356: eb7ff0500024 stmg %r7,%r15,80(%r15) Call Trace: ([<00000000001563e0>] dma_update_trans+0x90/0x228) [<00000000001565dc>] s390_dma_unmap_pages+0x64/0x160 [<00000000001567c2>] s390_dma_free+0x62/0x98 [<000003ff801310ce>] __genwqe_free_consistent+0x56/0x70 [genwqe_card] [<000003ff801316d0>] genwqe_free_sync_sgl+0xf8/0x160 [genwqe_card] [<000003ff8012bd6e>] ddcb_cmd_cleanup+0x86/0xa8 [genwqe_card] [<000003ff8012c1c0>] do_execute_ddcb+0x110/0x348 [genwqe_card] [<000003ff8012c914>] genwqe_ioctl+0x51c/0xc20 [genwqe_card] [<000000000032513a>] do_vfs_ioctl+0x3b2/0x518 [<0000000000325344>] SyS_ioctl+0xa4/0xb8 [<00000000007b86c6>] system_call+0xd6/0x264 [<000003ff9e8e520a>] 0x3ff9e8e520a Last Breaking-Event-Address: [<0000000000156342>] dma_update_cpu_trans+0x9a/0xa8 ---[ end trace 35996336235145c8 ]--- BUG: Bad page state in process jbd2/dasdb1-8 pfn:3215b page:000003d100c856c0 count:-1 mapcount:0 mapping: (null) index:0x0 flags: 0x3fffc0000000000() page dumped because: nonzero _count Signed-off-by: Gerald Schaefer <gerald.schaefer@de.ibm.com> Signed-off-by: Frank Haverkamp <haver@linux.vnet.ibm.com> Cc: stable <stable@vger.kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

commit a7a7aee upstream. When interrupting an application which was allocating DMAable memory, it was possible, that the DMA memory was deallocated twice, leading to the error symptoms below. Thanks to Gerald, who analyzed the problem and provided this patch. I agree with his analysis of the problem: ddcb_cmd_fixups() -> genwqe_alloc_sync_sgl() (fails in f/lpage, but sgl->sgl != NULL and f/lpage maybe also != NULL) -> ddcb_cmd_cleanup() -> genwqe_free_sync_sgl() (double free, because sgl->sgl != NULL and f/lpage maybe also != NULL) In this scenario we would have exactly the kind of double free that would explain the WARNING / Bad page state, and as expected it is caused by broken error handling (cleanup). Using the Ubuntu git source, tag Ubuntu-4.4.0-33.52, he was able to reproduce the "Bad page state" issue, and with the patch on top he could not reproduce it any more. ------------[ cut here ]------------ WARNING: at /build/linux-o03cxz/linux-4.4.0/arch/s390/include/asm/pci_dma.h:141 Modules linked in: qeth_l2 ghash_s390 prng aes_s390 des_s390 des_generic sha512_s390 sha256_s390 sha1_s390 sha_common genwqe_card qeth crc_itu_t qdio ccwgroup vmur dm_multipath dasd_eckd_mod dasd_mod CPU: 2 PID: 3293 Comm: genwqe_gunzip Not tainted 4.4.0-33-generic #52-Ubuntu task: 0000000032c7e270 ti: 00000000324e4000 task.ti: 00000000324e4000 Krnl PSW : 0404c00180000000 0000000000156346 (dma_update_cpu_trans+0x9e/0xa8) R:0 T:1 IO:0 EX:0 Key:0 M:1 W:0 P:0 AS:3 CC:0 PM:0 EA:3 Krnl GPRS: 00000000324e7bcd 0000000000c3c34a 0000000027628298 000000003215b400 0000000000000400 0000000000001fff 0000000000000400 0000000116853000 07000000324e7b1e 0000000000000001 0000000000000001 0000000000000001 0000000000001000 0000000116854000 0000000000156402 00000000324e7a38 Krnl Code: 000000000015633a: 95001000 cli 0(%r1),0 000000000015633e: a774ffc3 brc 7,1562c4 #0000000000156342: a7f40001 brc 15,156344 >0000000000156346: 92011000 mvi 0(%r1),1 000000000015634a: a7f4ffbd brc 15,1562c4 000000000015634e: 0707 bcr 0,%r7 0000000000156350: c00400000000 brcl 0,156350 0000000000156356: eb7ff0500024 stmg %r7,%r15,80(%r15) Call Trace: ([<00000000001563e0>] dma_update_trans+0x90/0x228) [<00000000001565dc>] s390_dma_unmap_pages+0x64/0x160 [<00000000001567c2>] s390_dma_free+0x62/0x98 [<000003ff801310ce>] __genwqe_free_consistent+0x56/0x70 [genwqe_card] [<000003ff801316d0>] genwqe_free_sync_sgl+0xf8/0x160 [genwqe_card] [<000003ff8012bd6e>] ddcb_cmd_cleanup+0x86/0xa8 [genwqe_card] [<000003ff8012c1c0>] do_execute_ddcb+0x110/0x348 [genwqe_card] [<000003ff8012c914>] genwqe_ioctl+0x51c/0xc20 [genwqe_card] [<000000000032513a>] do_vfs_ioctl+0x3b2/0x518 [<0000000000325344>] SyS_ioctl+0xa4/0xb8 [<00000000007b86c6>] system_call+0xd6/0x264 [<000003ff9e8e520a>] 0x3ff9e8e520a Last Breaking-Event-Address: [<0000000000156342>] dma_update_cpu_trans+0x9a/0xa8 ---[ end trace 35996336235145c8 ]--- BUG: Bad page state in process jbd2/dasdb1-8 pfn:3215b page:000003d100c856c0 count:-1 mapcount:0 mapping: (null) index:0x0 flags: 0x3fffc0000000000() page dumped because: nonzero _count Signed-off-by: Gerald Schaefer <gerald.schaefer@de.ibm.com> Signed-off-by: Frank Haverkamp <haver@linux.vnet.ibm.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Increase kasan instrumented kernel stack size from 32k to 64k. Other architectures seems to get away with just doubling kernel stack size under kasan, but on s390 this appears to be not enough due to bigger frame size. The particular pain point is kasan inlined checks (CONFIG_KASAN_INLINE vs CONFIG_KASAN_OUTLINE). With inlined checks one particular case hitting stack overflow is fs sync on xfs filesystem: #0 [9a0681e8] 704 bytes check_usage at 34b1fc #1 [9a0684a8] 432 bytes check_usage at 34c710 #2 [9a068658] 1048 bytes validate_chain at 35044a #3 [9a068a70] 312 bytes __lock_acquire at 3559fe #4 [9a068ba8] 440 bytes lock_acquire at 3576ee #5 [9a068d60] 104 bytes _raw_spin_lock at 21b44e0 #6 [9a068dc8] 1992 bytes enqueue_entity at 2dbf72 #7 [9a069590] 1496 bytes enqueue_task_fair at 2df5f0 #8 [9a069b68] 64 bytes ttwu_do_activate at 28f438 #9 [9a069ba8] 552 bytes try_to_wake_up at 298c4c #10 [9a069dd0] 168 bytes wake_up_worker at 23f97c #11 [9a069e78] 200 bytes insert_work at 23fc2e #12 [9a069f40] 648 bytes __queue_work at 2487c0 #13 [9a06a1c8] 200 bytes __queue_delayed_work at 24db28 #14 [9a06a290] 248 bytes mod_delayed_work_on at 24de84 #15 [9a06a388] 24 bytes kblockd_mod_delayed_work_on at 153e2a0 #16 [9a06a3a0] 288 bytes __blk_mq_delay_run_hw_queue at 158168c #17 [9a06a4c0] 192 bytes blk_mq_run_hw_queue at 1581a3c #18 [9a06a580] 184 bytes blk_mq_sched_insert_requests at 15a2192 #19 [9a06a638] 1024 bytes blk_mq_flush_plug_list at 1590f3a #20 [9a06aa38] 704 bytes blk_flush_plug_list at 1555028 #21 [9a06acf8] 320 bytes schedule at 219e476 #22 [9a06ae38] 760 bytes schedule_timeout at 21b0aac #23 [9a06b130] 408 bytes wait_for_common at 21a1706 #24 [9a06b2c8] 360 bytes xfs_buf_iowait at fa1540 #25 [9a06b430] 256 bytes __xfs_buf_submit at fadae6 #26 [9a06b530] 264 bytes xfs_buf_read_map at fae3f6 #27 [9a06b638] 656 bytes xfs_trans_read_buf_map at 10ac9a8 #28 [9a06b8c8] 304 bytes xfs_btree_kill_root at e72426 #29 [9a06b9f8] 288 bytes xfs_btree_lookup_get_block at e7bc5e #30 [9a06bb18] 624 bytes xfs_btree_lookup at e7e1a6 #31 [9a06bd88] 2664 bytes xfs_alloc_ag_vextent_near at dfa070 #32 [9a06c7f0] 144 bytes xfs_alloc_ag_vextent at dff3ca #33 [9a06c880] 1128 bytes xfs_alloc_vextent at e05fce #34 [9a06cce8] 584 bytes xfs_bmap_btalloc at e58342 #35 [9a06cf30] 1336 bytes xfs_bmapi_write at e618de #36 [9a06d468] 776 bytes xfs_iomap_write_allocate at ff678e #37 [9a06d770] 720 bytes xfs_map_blocks at f82af8 #38 [9a06da40] 928 bytes xfs_writepage_map at f83cd6 #39 [9a06dde0] 320 bytes xfs_do_writepage at f85872 #40 [9a06df20] 1320 bytes write_cache_pages at 73dfe8 #41 [9a06e448] 208 bytes xfs_vm_writepages at f7f892 #42 [9a06e518] 88 bytes do_writepages at 73fe6a #43 [9a06e570] 872 bytes __writeback_single_inode at a20cb6 #44 [9a06e8d8] 664 bytes writeback_sb_inodes at a23be2 #45 [9a06eb70] 296 bytes __writeback_inodes_wb at a242e0 #46 [9a06ec98] 928 bytes wb_writeback at a2500e #47 [9a06f038] 848 bytes wb_do_writeback at a260ae #48 [9a06f388] 536 bytes wb_workfn at a28228 #49 [9a06f5a0] 1088 bytes process_one_work at 24a234 #50 [9a06f9e0] 1120 bytes worker_thread at 24ba26 #51 [9a06fe40] 104 bytes kthread at 26545a #52 [9a06fea8] kernel_thread_starter at 21b6b62 To be able to increase the stack size to 64k reuse LLILL instruction in __switch_to function to load 64k - STACK_FRAME_OVERHEAD - __PT_SIZE (65192) value as unsigned. Reported-by: Benjamin Block <bblock@linux.ibm.com> Reviewed-by: Heiko Carstens <heiko.carstens@de.ibm.com> Signed-off-by: Vasily Gorbik <gor@linux.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>

commit 64b9cea upstream. Syzkaller was able to trigger the following splat again: WARNING: CPU: 1 PID: 12512 at net/mptcp/protocol.c:761 mptcp_reset_timer+0x12a/0x160 net/mptcp/protocol.c:761 Modules linked in: CPU: 1 PID: 12512 Comm: kworker/1:6 Not tainted 5.10.0-rc6 #52 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.13.0-0-gf21b5a4aeb02-prebuilt.qemu.org 04/01/2014 Workqueue: events mptcp_worker RIP: 0010:mptcp_reset_timer+0x12a/0x160 net/mptcp/protocol.c:761 Code: e8 4b 0c ad ff e8 56 21 88 fe 48 b8 00 00 00 00 00 fc ff df 48 c7 04 03 00 00 00 00 48 83 c4 40 5b 5d 41 5c c3 e8 36 21 88 fe <0f> 0b 41 bc c8 00 00 00 eb 98 e8 e7 b1 af fe e9 30 ff ff ff 48 c7 RSP: 0018:ffffc900018c7c68 EFLAGS: 00010293 RAX: ffff888108cb1c80 RBX: 1ffff92000318f8d RCX: ffffffff82ad0307 RDX: 0000000000000000 RSI: ffffffff82ad036a RDI: 0000000000000007 RBP: ffff888113e2d000 R08: ffff888108cb1c80 R09: ffffed10227c5ab7 R10: ffff888113e2d5b7 R11: ffffed10227c5ab6 R12: 0000000000000000 R13: ffff88801f100000 R14: ffff888113e2d5b0 R15: 0000000000000001 FS: 0000000000000000(0000) GS:ffff88811b500000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00007fd76a874ef8 CR3: 000000001689c005 CR4: 0000000000170ee0 Call Trace: mptcp_worker+0xaa4/0x1560 net/mptcp/protocol.c:2334 process_one_work+0x8d3/0x1200 kernel/workqueue.c:2272 worker_thread+0x9c/0x1090 kernel/workqueue.c:2418 kthread+0x303/0x410 kernel/kthread.c:292 ret_from_fork+0x22/0x30 arch/x86/entry/entry_64.S:296 The mptcp_worker tries to update the MPTCP retransmission timer even if such timer is not currently scheduled. The mptcp_rtx_head() return value is bogus: we can have enqueued data not yet transmitted. The above may additionally cause spurious, unneeded MPTCP-level retransmissions. Fix the issue adding an explicit clearing of the rtx queue before trying to retransmit and checking for unacked data. Additionally drop an unneeded timer stop call and the unused mptcp_rtx_tail() helper. Reported-by: Christoph Paasch <cpaasch@apple.com> Fixes: 6e628cd ("mptcp: use mptcp release_cb for delayed tasks") Signed-off-by: Paolo Abeni <pabeni@redhat.com> Signed-off-by: Mat Martineau <mathew.j.martineau@linux.intel.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

[ Upstream commit d5027ca ] Ritesh reported a bug [1] against UML, noting that it crashed on startup. The backtrace shows the following (heavily redacted): (gdb) bt ... #26 0x0000000060015b5d in sem_init () at ipc/sem.c:268 #27 0x00007f89906d92f7 in ?? () from /lib/x86_64-linux-gnu/libcom_err.so.2 #28 0x00007f8990ab8fb2 in call_init (...) at dl-init.c:72 ... #40 0x00007f89909bf3a6 in nss_load_library (...) at nsswitch.c:359 ... #44 0x00007f8990895e35 in _nss_compat_getgrnam_r (...) at nss_compat/compat-grp.c:486 #45 0x00007f8990968b85 in __getgrnam_r [...] #46 0x00007f89909d6b77 in grantpt [...] #47 0x00007f8990a9394e in __GI_openpty [...] #48 0x00000000604a1f65 in openpty_cb (...) at arch/um/os-Linux/sigio.c:407 #49 0x00000000604a58d0 in start_idle_thread (...) at arch/um/os-Linux/skas/process.c:598 #50 0x0000000060004a3d in start_uml () at arch/um/kernel/skas/process.c:45 #51 0x00000000600047b2 in linux_main (...) at arch/um/kernel/um_arch.c:334 #52 0x000000006000574f in main (...) at arch/um/os-Linux/main.c:144 indicating that the UML function openpty_cb() calls openpty(), which internally calls __getgrnam_r(), which causes the nsswitch machinery to get started. This loads, through lots of indirection that I snipped, the libcom_err.so.2 library, which (in an unknown function, "??") calls sem_init(). Now, of course it wants to get libpthread's sem_init(), since it's linked against libpthread. However, the dynamic linker looks up that symbol against the binary first, and gets the kernel's sem_init(). Hajime Tazaki noted that "objcopy -L" can localize a symbol, so the dynamic linker wouldn't do the lookup this way. I tried, but for some reason that didn't seem to work. Doing the same thing in the linker script instead does seem to work, though I cannot entirely explain - it *also* works if I just add "VERSION { { global: *; }; }" instead, indicating that something else is happening that I don't really understand. It may be that explicitly doing that marks them with some kind of empty version, and that's different from the default. Explicitly marking them with a version breaks kallsyms, so that doesn't seem to be possible. Marking all the symbols as local seems correct, and does seem to address the issue, so do that. Also do it for static link, nsswitch libraries could still be loaded there. [1] https://bugs.debian.org/983379 Reported-by: Ritesh Raj Sarraf <rrs@debian.org> Signed-off-by: Johannes Berg <johannes.berg@intel.com> Acked-By: Anton Ivanov <anton.ivanov@cambridgegreys.com> Tested-By: Ritesh Raj Sarraf <rrs@debian.org> Signed-off-by: Richard Weinberger <richard@nod.at> Signed-off-by: Sasha Levin <sashal@kernel.org>

The inline assembly for arm64's cmpxchg_double*() implementations use a +Q constraint to hazard against other accesses to the memory location being exchanged. However, the pointer passed to the constraint is a pointer to unsigned long, and thus the hazard only applies to the first 8 bytes of the location. GCC can take advantage of this, assuming that other portions of the location are unchanged, leading to a number of potential problems. This is similar to what we fixed back in commit: fee960b ("arm64: xchg: hazard against entire exchange variable") ... but we forgot to adjust cmpxchg_double*() similarly at the same time. The same problem applies, as demonstrated with the following test: | struct big { | u64 lo, hi; | } __aligned(128); | | unsigned long foo(struct big *b) | { | u64 hi_old, hi_new; | | hi_old = b->hi; | cmpxchg_double_local(&b->lo, &b->hi, 0x12, 0x34, 0x56, 0x78); | hi_new = b->hi; | | return hi_old ^ hi_new; | } ... which GCC 12.1.0 compiles as: | 0000000000000000 <foo>: | 0: d503233f paciasp | 4: aa0003e4 mov x4, x0 | 8: 1400000e b 40 <foo+0x40> | c: d2800240 mov x0, #0x12 // #18 | 10: d2800681 mov x1, #0x34 // #52 | 14: aa0003e5 mov x5, x0 | 18: aa0103e6 mov x6, x1 | 1c: d2800ac2 mov x2, #0x56 // #86 | 20: d2800f03 mov x3, #0x78 // #120 | 24: 48207c82 casp x0, x1, x2, x3, [x4] | 28: ca050000 eor x0, x0, x5 | 2c: ca060021 eor x1, x1, x6 | 30: aa010000 orr x0, x0, x1 | 34: d2800000 mov x0, #0x0 // #0 <--- BANG | 38: d50323bf autiasp | 3c: d65f03c0 ret | 40: d2800240 mov x0, #0x12 // #18 | 44: d2800681 mov x1, #0x34 // #52 | 48: d2800ac2 mov x2, #0x56 // #86 | 4c: d2800f03 mov x3, #0x78 // #120 | 50: f9800091 prfm pstl1strm, [x4] | 54: c87f1885 ldxp x5, x6, [x4] | 58: ca0000a5 eor x5, x5, x0 | 5c: ca0100c6 eor x6, x6, x1 | 60: aa0600a6 orr x6, x5, x6 | 64: b5000066 cbnz x6, 70 <foo+0x70> | 68: c8250c82 stxp w5, x2, x3, [x4] | 6c: 35ffff45 cbnz w5, 54 <foo+0x54> | 70: d2800000 mov x0, #0x0 // #0 <--- BANG | 74: d50323bf autiasp | 78: d65f03c0 ret Notice that at the lines with "BANG" comments, GCC has assumed that the higher 8 bytes are unchanged by the cmpxchg_double() call, and that `hi_old ^ hi_new` can be reduced to a constant zero, for both LSE and LL/SC versions of cmpxchg_double(). This patch fixes the issue by passing a pointer to __uint128_t into the +Q constraint, ensuring that the compiler hazards against the entire 16 bytes being modified. With this change, GCC 12.1.0 compiles the above test as: | 0000000000000000 <foo>: | 0: f9400407 ldr x7, [x0, #8] | 4: d503233f paciasp | 8: aa0003e4 mov x4, x0 | c: 1400000f b 48 <foo+0x48> | 10: d2800240 mov x0, #0x12 // #18 | 14: d2800681 mov x1, #0x34 // #52 | 18: aa0003e5 mov x5, x0 | 1c: aa0103e6 mov x6, x1 | 20: d2800ac2 mov x2, #0x56 // #86 | 24: d2800f03 mov x3, #0x78 // #120 | 28: 48207c82 casp x0, x1, x2, x3, [x4] | 2c: ca050000 eor x0, x0, x5 | 30: ca060021 eor x1, x1, x6 | 34: aa010000 orr x0, x0, x1 | 38: f9400480 ldr x0, [x4, #8] | 3c: d50323bf autiasp | 40: ca0000e0 eor x0, x7, x0 | 44: d65f03c0 ret | 48: d2800240 mov x0, #0x12 // #18 | 4c: d2800681 mov x1, #0x34 // #52 | 50: d2800ac2 mov x2, #0x56 // #86 | 54: d2800f03 mov x3, #0x78 // #120 | 58: f9800091 prfm pstl1strm, [x4] | 5c: c87f1885 ldxp x5, x6, [x4] | 60: ca0000a5 eor x5, x5, x0 | 64: ca0100c6 eor x6, x6, x1 | 68: aa0600a6 orr x6, x5, x6 | 6c: b5000066 cbnz x6, 78 <foo+0x78> | 70: c8250c82 stxp w5, x2, x3, [x4] | 74: 35ffff45 cbnz w5, 5c <foo+0x5c> | 78: f9400480 ldr x0, [x4, #8] | 7c: d50323bf autiasp | 80: ca0000e0 eor x0, x7, x0 | 84: d65f03c0 ret ... sampling the high 8 bytes before and after the cmpxchg, and performing an EOR, as we'd expect. For backporting, I've tested this atop linux-4.9.y with GCC 5.5.0. Note that linux-4.9.y is oldest currently supported stable release, and mandates GCC 5.1+. Unfortunately I couldn't get a GCC 5.1 binary to run on my machines due to library incompatibilities. I've also used a standalone test to check that we can use a __uint128_t pointer in a +Q constraint at least as far back as GCC 4.8.5 and LLVM 3.9.1. Fixes: 5284e1b ("arm64: xchg: Implement cmpxchg_double") Fixes: e9a4b79 ("arm64: cmpxchg_dbl: patch in lse instructions when supported by the CPU") Reported-by: Boqun Feng <boqun.feng@gmail.com> Link: https://lore.kernel.org/lkml/Y6DEfQXymYVgL3oJ@boqun-archlinux/ Reported-by: Peter Zijlstra <peterz@infradead.org> Link: https://lore.kernel.org/lkml/Y6GXoO4qmH9OIZ5Q@hirez.programming.kicks-ass.net/ Signed-off-by: Mark Rutland <mark.rutland@arm.com> Cc: stable@vger.kernel.org Cc: Arnd Bergmann <arnd@arndb.de> Cc: Catalin Marinas <catalin.marinas@arm.com> Cc: Steve Capper <steve.capper@arm.com> Cc: Will Deacon <will@kernel.org> Link: https://lore.kernel.org/r/20230104151626.3262137-1-mark.rutland@arm.com Signed-off-by: Will Deacon <will@kernel.org>

commit 031af50 upstream. The inline assembly for arm64's cmpxchg_double*() implementations use a +Q constraint to hazard against other accesses to the memory location being exchanged. However, the pointer passed to the constraint is a pointer to unsigned long, and thus the hazard only applies to the first 8 bytes of the location. GCC can take advantage of this, assuming that other portions of the location are unchanged, leading to a number of potential problems. This is similar to what we fixed back in commit: fee960b ("arm64: xchg: hazard against entire exchange variable") ... but we forgot to adjust cmpxchg_double*() similarly at the same time. The same problem applies, as demonstrated with the following test: | struct big { | u64 lo, hi; | } __aligned(128); | | unsigned long foo(struct big *b) | { | u64 hi_old, hi_new; | | hi_old = b->hi; | cmpxchg_double_local(&b->lo, &b->hi, 0x12, 0x34, 0x56, 0x78); | hi_new = b->hi; | | return hi_old ^ hi_new; | } ... which GCC 12.1.0 compiles as: | 0000000000000000 <foo>: | 0: d503233f paciasp | 4: aa0003e4 mov x4, x0 | 8: 1400000e b 40 <foo+0x40> | c: d2800240 mov x0, #0x12 // #18 | 10: d2800681 mov x1, #0x34 // #52 | 14: aa0003e5 mov x5, x0 | 18: aa0103e6 mov x6, x1 | 1c: d2800ac2 mov x2, #0x56 // #86 | 20: d2800f03 mov x3, #0x78 // #120 | 24: 48207c82 casp x0, x1, x2, x3, [x4] | 28: ca050000 eor x0, x0, x5 | 2c: ca060021 eor x1, x1, x6 | 30: aa010000 orr x0, x0, x1 | 34: d2800000 mov x0, #0x0 // #0 <--- BANG | 38: d50323bf autiasp | 3c: d65f03c0 ret | 40: d2800240 mov x0, #0x12 // #18 | 44: d2800681 mov x1, #0x34 // #52 | 48: d2800ac2 mov x2, #0x56 // #86 | 4c: d2800f03 mov x3, #0x78 // #120 | 50: f9800091 prfm pstl1strm, [x4] | 54: c87f1885 ldxp x5, x6, [x4] | 58: ca0000a5 eor x5, x5, x0 | 5c: ca0100c6 eor x6, x6, x1 | 60: aa0600a6 orr x6, x5, x6 | 64: b5000066 cbnz x6, 70 <foo+0x70> | 68: c8250c82 stxp w5, x2, x3, [x4] | 6c: 35ffff45 cbnz w5, 54 <foo+0x54> | 70: d2800000 mov x0, #0x0 // #0 <--- BANG | 74: d50323bf autiasp | 78: d65f03c0 ret Notice that at the lines with "BANG" comments, GCC has assumed that the higher 8 bytes are unchanged by the cmpxchg_double() call, and that `hi_old ^ hi_new` can be reduced to a constant zero, for both LSE and LL/SC versions of cmpxchg_double(). This patch fixes the issue by passing a pointer to __uint128_t into the +Q constraint, ensuring that the compiler hazards against the entire 16 bytes being modified. With this change, GCC 12.1.0 compiles the above test as: | 0000000000000000 <foo>: | 0: f9400407 ldr x7, [x0, #8] | 4: d503233f paciasp | 8: aa0003e4 mov x4, x0 | c: 1400000f b 48 <foo+0x48> | 10: d2800240 mov x0, #0x12 // #18 | 14: d2800681 mov x1, #0x34 // #52 | 18: aa0003e5 mov x5, x0 | 1c: aa0103e6 mov x6, x1 | 20: d2800ac2 mov x2, #0x56 // #86 | 24: d2800f03 mov x3, #0x78 // #120 | 28: 48207c82 casp x0, x1, x2, x3, [x4] | 2c: ca050000 eor x0, x0, x5 | 30: ca060021 eor x1, x1, x6 | 34: aa010000 orr x0, x0, x1 | 38: f9400480 ldr x0, [x4, #8] | 3c: d50323bf autiasp | 40: ca0000e0 eor x0, x7, x0 | 44: d65f03c0 ret | 48: d2800240 mov x0, #0x12 // #18 | 4c: d2800681 mov x1, #0x34 // #52 | 50: d2800ac2 mov x2, #0x56 // #86 | 54: d2800f03 mov x3, #0x78 // #120 | 58: f9800091 prfm pstl1strm, [x4] | 5c: c87f1885 ldxp x5, x6, [x4] | 60: ca0000a5 eor x5, x5, x0 | 64: ca0100c6 eor x6, x6, x1 | 68: aa0600a6 orr x6, x5, x6 | 6c: b5000066 cbnz x6, 78 <foo+0x78> | 70: c8250c82 stxp w5, x2, x3, [x4] | 74: 35ffff45 cbnz w5, 5c <foo+0x5c> | 78: f9400480 ldr x0, [x4, #8] | 7c: d50323bf autiasp | 80: ca0000e0 eor x0, x7, x0 | 84: d65f03c0 ret ... sampling the high 8 bytes before and after the cmpxchg, and performing an EOR, as we'd expect. For backporting, I've tested this atop linux-4.9.y with GCC 5.5.0. Note that linux-4.9.y is oldest currently supported stable release, and mandates GCC 5.1+. Unfortunately I couldn't get a GCC 5.1 binary to run on my machines due to library incompatibilities. I've also used a standalone test to check that we can use a __uint128_t pointer in a +Q constraint at least as far back as GCC 4.8.5 and LLVM 3.9.1. Fixes: 5284e1b ("arm64: xchg: Implement cmpxchg_double") Fixes: e9a4b79 ("arm64: cmpxchg_dbl: patch in lse instructions when supported by the CPU") Reported-by: Boqun Feng <boqun.feng@gmail.com> Link: https://lore.kernel.org/lkml/Y6DEfQXymYVgL3oJ@boqun-archlinux/ Reported-by: Peter Zijlstra <peterz@infradead.org> Link: https://lore.kernel.org/lkml/Y6GXoO4qmH9OIZ5Q@hirez.programming.kicks-ass.net/ Signed-off-by: Mark Rutland <mark.rutland@arm.com> Cc: stable@vger.kernel.org Cc: Arnd Bergmann <arnd@arndb.de> Cc: Catalin Marinas <catalin.marinas@arm.com> Cc: Steve Capper <steve.capper@arm.com> Cc: Will Deacon <will@kernel.org> Link: https://lore.kernel.org/r/20230104151626.3262137-1-mark.rutland@arm.com Signed-off-by: Will Deacon <will@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

[ Upstream commit 031af50 ] The inline assembly for arm64's cmpxchg_double*() implementations use a +Q constraint to hazard against other accesses to the memory location being exchanged. However, the pointer passed to the constraint is a pointer to unsigned long, and thus the hazard only applies to the first 8 bytes of the location. GCC can take advantage of this, assuming that other portions of the location are unchanged, leading to a number of potential problems. This is similar to what we fixed back in commit: fee960b ("arm64: xchg: hazard against entire exchange variable") ... but we forgot to adjust cmpxchg_double*() similarly at the same time. The same problem applies, as demonstrated with the following test: | struct big { | u64 lo, hi; | } __aligned(128); | | unsigned long foo(struct big *b) | { | u64 hi_old, hi_new; | | hi_old = b->hi; | cmpxchg_double_local(&b->lo, &b->hi, 0x12, 0x34, 0x56, 0x78); | hi_new = b->hi; | | return hi_old ^ hi_new; | } ... which GCC 12.1.0 compiles as: | 0000000000000000 <foo>: | 0: d503233f paciasp | 4: aa0003e4 mov x4, x0 | 8: 1400000e b 40 <foo+0x40> | c: d2800240 mov x0, #0x12 // #18 | 10: d2800681 mov x1, #0x34 // #52 | 14: aa0003e5 mov x5, x0 | 18: aa0103e6 mov x6, x1 | 1c: d2800ac2 mov x2, #0x56 // #86 | 20: d2800f03 mov x3, #0x78 // #120 | 24: 48207c82 casp x0, x1, x2, x3, [x4] | 28: ca050000 eor x0, x0, x5 | 2c: ca060021 eor x1, x1, x6 | 30: aa010000 orr x0, x0, x1 | 34: d2800000 mov x0, #0x0 // #0 <--- BANG | 38: d50323bf autiasp | 3c: d65f03c0 ret | 40: d2800240 mov x0, #0x12 // #18 | 44: d2800681 mov x1, #0x34 // #52 | 48: d2800ac2 mov x2, #0x56 // #86 | 4c: d2800f03 mov x3, #0x78 // #120 | 50: f9800091 prfm pstl1strm, [x4] | 54: c87f1885 ldxp x5, x6, [x4] | 58: ca0000a5 eor x5, x5, x0 | 5c: ca0100c6 eor x6, x6, x1 | 60: aa0600a6 orr x6, x5, x6 | 64: b5000066 cbnz x6, 70 <foo+0x70> | 68: c8250c82 stxp w5, x2, x3, [x4] | 6c: 35ffff45 cbnz w5, 54 <foo+0x54> | 70: d2800000 mov x0, #0x0 // #0 <--- BANG | 74: d50323bf autiasp | 78: d65f03c0 ret Notice that at the lines with "BANG" comments, GCC has assumed that the higher 8 bytes are unchanged by the cmpxchg_double() call, and that `hi_old ^ hi_new` can be reduced to a constant zero, for both LSE and LL/SC versions of cmpxchg_double(). This patch fixes the issue by passing a pointer to __uint128_t into the +Q constraint, ensuring that the compiler hazards against the entire 16 bytes being modified. With this change, GCC 12.1.0 compiles the above test as: | 0000000000000000 <foo>: | 0: f9400407 ldr x7, [x0, #8] | 4: d503233f paciasp | 8: aa0003e4 mov x4, x0 | c: 1400000f b 48 <foo+0x48> | 10: d2800240 mov x0, #0x12 // #18 | 14: d2800681 mov x1, #0x34 // #52 | 18: aa0003e5 mov x5, x0 | 1c: aa0103e6 mov x6, x1 | 20: d2800ac2 mov x2, #0x56 // #86 | 24: d2800f03 mov x3, #0x78 // #120 | 28: 48207c82 casp x0, x1, x2, x3, [x4] | 2c: ca050000 eor x0, x0, x5 | 30: ca060021 eor x1, x1, x6 | 34: aa010000 orr x0, x0, x1 | 38: f9400480 ldr x0, [x4, #8] | 3c: d50323bf autiasp | 40: ca0000e0 eor x0, x7, x0 | 44: d65f03c0 ret | 48: d2800240 mov x0, #0x12 // #18 | 4c: d2800681 mov x1, #0x34 // #52 | 50: d2800ac2 mov x2, #0x56 // #86 | 54: d2800f03 mov x3, #0x78 // #120 | 58: f9800091 prfm pstl1strm, [x4] | 5c: c87f1885 ldxp x5, x6, [x4] | 60: ca0000a5 eor x5, x5, x0 | 64: ca0100c6 eor x6, x6, x1 | 68: aa0600a6 orr x6, x5, x6 | 6c: b5000066 cbnz x6, 78 <foo+0x78> | 70: c8250c82 stxp w5, x2, x3, [x4] | 74: 35ffff45 cbnz w5, 5c <foo+0x5c> | 78: f9400480 ldr x0, [x4, #8] | 7c: d50323bf autiasp | 80: ca0000e0 eor x0, x7, x0 | 84: d65f03c0 ret ... sampling the high 8 bytes before and after the cmpxchg, and performing an EOR, as we'd expect. For backporting, I've tested this atop linux-4.9.y with GCC 5.5.0. Note that linux-4.9.y is oldest currently supported stable release, and mandates GCC 5.1+. Unfortunately I couldn't get a GCC 5.1 binary to run on my machines due to library incompatibilities. I've also used a standalone test to check that we can use a __uint128_t pointer in a +Q constraint at least as far back as GCC 4.8.5 and LLVM 3.9.1. Fixes: 5284e1b ("arm64: xchg: Implement cmpxchg_double") Fixes: e9a4b79 ("arm64: cmpxchg_dbl: patch in lse instructions when supported by the CPU") Reported-by: Boqun Feng <boqun.feng@gmail.com> Link: https://lore.kernel.org/lkml/Y6DEfQXymYVgL3oJ@boqun-archlinux/ Reported-by: Peter Zijlstra <peterz@infradead.org> Link: https://lore.kernel.org/lkml/Y6GXoO4qmH9OIZ5Q@hirez.programming.kicks-ass.net/ Signed-off-by: Mark Rutland <mark.rutland@arm.com> Cc: stable@vger.kernel.org Cc: Arnd Bergmann <arnd@arndb.de> Cc: Catalin Marinas <catalin.marinas@arm.com> Cc: Steve Capper <steve.capper@arm.com> Cc: Will Deacon <will@kernel.org> Link: https://lore.kernel.org/r/20230104151626.3262137-1-mark.rutland@arm.com Signed-off-by: Will Deacon <will@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>

central: Pin timer callbacks to central CPU

ghollingworth closed this as completed Jul 17, 2013

0lxb pushed a commit to 0lxb/rpi_linux that referenced this issue Jan 30, 2024

Merge pull request raspberrypi#52 from sched-ext/central_cpu_pin

b41ddc3

central: Pin timer callbacks to central CPU

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Crash when use of the sd card + network card #52

Crash when use of the sd card + network card #52

alphaonex86 commented Jul 11, 2012

popcornmix commented Jul 11, 2012

alphaonex86 commented Jul 11, 2012

popcornmix commented Jul 11, 2012

alphaonex86 commented Jul 12, 2012

alphaonex86 commented Jul 12, 2012

popcornmix commented Jul 12, 2012

alphaonex86 commented Jul 12, 2012

alphaonex86 commented Jul 12, 2012

alphaonex86 commented Jul 12, 2012

alphaonex86 commented Jul 12, 2012

alphaonex86 commented Jul 12, 2012

alphaonex86 commented Jul 14, 2012

alphaonex86 commented Jul 16, 2012

alphaonex86 commented Jul 16, 2012

alphaonex86 commented Jul 25, 2012

alphaonex86 commented Aug 20, 2012

popcornmix commented Aug 20, 2012

alphaonex86 commented Aug 26, 2012

ashleyw commented Aug 27, 2012

alphaonex86 commented Aug 27, 2012

alphaonex86 commented Oct 10, 2012

popcornmix commented Oct 10, 2012

alphaonex86 commented Oct 10, 2012

alphaonex86 commented Oct 11, 2012

alphaonex86 commented Oct 19, 2012

alphaonex86 commented Nov 6, 2012

ghollingworth commented Nov 7, 2012

alphaonex86 commented Nov 7, 2012

alphaonex86 commented Nov 10, 2012

ghollingworth commented Nov 10, 2012

alphaonex86 commented Nov 10, 2012

ghollingworth commented Jul 17, 2013

Crash when use of the sd card + network card #52

Crash when use of the sd card + network card #52

Comments

alphaonex86 commented Jul 11, 2012

popcornmix commented Jul 11, 2012

alphaonex86 commented Jul 11, 2012

popcornmix commented Jul 11, 2012

alphaonex86 commented Jul 12, 2012

alphaonex86 commented Jul 12, 2012

popcornmix commented Jul 12, 2012

alphaonex86 commented Jul 12, 2012

alphaonex86 commented Jul 12, 2012

alphaonex86 commented Jul 12, 2012

alphaonex86 commented Jul 12, 2012

alphaonex86 commented Jul 12, 2012

alphaonex86 commented Jul 14, 2012

alphaonex86 commented Jul 16, 2012

alphaonex86 commented Jul 16, 2012

alphaonex86 commented Jul 25, 2012

alphaonex86 commented Aug 20, 2012

popcornmix commented Aug 20, 2012

alphaonex86 commented Aug 26, 2012

ashleyw commented Aug 27, 2012

alphaonex86 commented Aug 27, 2012

alphaonex86 commented Oct 10, 2012

popcornmix commented Oct 10, 2012

alphaonex86 commented Oct 10, 2012

alphaonex86 commented Oct 11, 2012

alphaonex86 commented Oct 19, 2012

alphaonex86 commented Nov 6, 2012

ghollingworth commented Nov 7, 2012

alphaonex86 commented Nov 7, 2012

alphaonex86 commented Nov 10, 2012

ghollingworth commented Nov 10, 2012

alphaonex86 commented Nov 10, 2012

ghollingworth commented Jul 17, 2013