Fix 'sleeping function called from invalid context' warning in sysrq generated crash.
PATCH 1/1 Fix 'sleeping function called from invalid context' warning in sysrq generated crash.
Commit 984d74a72076a1 ("sysrq: rcu-ify __handle_sysrq") replaced spin_lock_irqsave() calls with rcu_read_lock() calls in sysrq. Since rcu_read_lock() does not disable preemption, faulthandler_disabled() in __do_page_fault() in x86/fault.c returns false. When the code later calls might_sleep() in the pagefault handler, we get the following warning:
BUG: sleeping function called from invalid context at ../arch/x86/mm/fault.c:1187
in_atomic(): 0, irqs_disabled(): 0, pid: 4706, name: bash Preemption disabled at:[<ffffffff81484339>] printk+0x48/0x4a
To fix this, we release the RCU read lock before we crash.
Tested this patch on linux 3.18 by booting off one of our boards.
--- drivers/tty/sysrq.c | 6 ++++++ 1 file changed, 6 insertions(+)
drivers/tty/sysrq.c +6 -0
@@ -133,6 +133,12 @@ static void sysrq_handle_crash(int key) {char *killer =NULL ; +/* we need to release the RCU read lock here, +otherwise we get an annoying +'BUG: sleeping function called from invalid context' +complaint from the kernel before the panic. +*/ +rcu_read_unlock (); panic_on_oops =1 ;/* force panic */ wmb (); *killer =1 ;
Signature · 1 line
1.8.1.4
Quoted 12 lines · Ani Sinha
Commit
984d74a72076a1("sysrq: rcu-ify__handle_sysrq") replacedspin_lock_irqsave()calls withrcu_read_lock()calls in sysrq. Sincercu_read_lock()does not disable preemption,faulthandler_disabled()in__do_page_fault()inx86/fault.creturns false. When the code later callsmight_sleep()in the pagefault handler, we get the following warning:BUG: sleeping function called from invalid context at ../arch/x86/mm/fault.c:1187
in_atomic(): 0,irqs_disabled(): 0, pid: 4706, name: bash Preemption disabled at:[<ffffffff81484339>]printk+0x48/0x4a
Quoted 7 lines
To fix this, we release the RCU read lock before we crash.
Tested this patch on linux 3.18 by booting off one of our boards.
Fixes 984d74a72076a1 ("sysrq: rcu-ify __handle_sysrq")Signed-off-by Ani Sinha <[email protected]>
I queued this with Rik's Signed-off-by, and fixed some checkpatch.pl errors. Please run scripts/checkpatch.pl on your patches in the future.
Please see below for the result.
Rik, did you test this as well? If so, may I also have your Tested-by?
Thanx, Paul
------------------------------------------------------------------------
commit c95a158356397844a5a6deb0bd58758084f891df
Author: Ani Sinha <[email protected]> Date: Thu Dec 17 17:15:10 2015 -0800
sysrq: Fix warning in sysrq generated crash.
Commit 984d74a72076a1 ("sysrq: rcu-ify __handle_sysrq") replaced spin_lock_irqsave() calls with rcu_read_lock() calls in sysrq. Since rcu_read_lock() does not disable preemption, faulthandler_disabled() in __do_page_fault() in x86/fault.c returns false. When the code later calls might_sleep() in the pagefault handler, we get the following warning:
BUG: sleeping function called from invalid context at ../arch/x86/mm/fault.c:1187 in_atomic(): 0, irqs_disabled(): 0, pid: 4706, name: bash Preemption disabled at:[<ffffffff81484339>] printk+0x48/0x4a
To fix this, we release the RCU read lock before we crash.
Tested this patch on linux 3.18 by booting off one of our boards.
drivers/tty/sysrq.c +6 -0
@@ -133,6 +133,12 @@ static void sysrq_handle_crash(int key) {char *killer =NULL ; +/* we need to release the RCU read lock here, +* otherwise we get an annoying +* 'BUG: sleeping function called from invalid context' +* complaint from the kernel before the panic. +*/ +rcu_read_unlock (); panic_on_oops =1 ;/* force panic */ wmb (); *killer =1 ;
Quoted 17 lines · Paul E. McKenney
Quoted 16 lines · Ani Sinha
Commit
984d74a72076a1("sysrq: rcu-ify__handle_sysrq") replacedspin_lock_irqsave()calls withrcu_read_lock()calls in sysrq. Sincercu_read_lock()does not disable preemption,faulthandler_disabled()in__do_page_fault()inx86/fault.creturns false. When the code later callsmight_sleep()in the pagefault handler, we get the following warning:BUG: sleeping function called from invalid context at ../arch/x86/mm/fault.c:1187
in_atomic(): 0,irqs_disabled(): 0, pid: 4706, name: bash Preemption disabled at:[<ffffffff81484339>]printk+0x48/0x4aTo fix this, we release the RCU read lock before we crash.
Tested this patch on linux 3.18 by booting off one of our boards.
Quoted 8 lines
Quoted 3 lines
Fixes 984d74a72076a1 ("sysrq: rcu-ify __handle_sysrq")Signed-off-by Ani Sinha <[email protected]>I queued this with Rik's Signed-off-by, and fixed some checkpatch.pl errors. Please run
scripts/checkpatch.plon your patches in the future.Please see below for the result.
Yes, sorry. I forgot to run the script before sending Thanks for fixing my comment and the trailing whitespace.
Quoted 51 lines
Rik, did you test this as well? If so, may I also have your Tested-by?
Thanx, Paul
------------------------------------------------------------------------
commit
c95a158356397844a5a6deb0bd58758084f891dfAuthor: Ani Sinha <[email protected]> Date: Thu Dec 17 17:15:10 2015 -0800sysrq: Fix warning in sysrq generated crash.
Commit
984d74a72076a1("sysrq: rcu-ify__handle_sysrq") replacedspin_lock_irqsave()calls withrcu_read_lock()calls in sysrq. Sincercu_read_lock()does not disable preemption,faulthandler_disabled()in__do_page_fault()inx86/fault.creturns false. When the code later callsmight_sleep()in the pagefault handler, we get the following warning:BUG: sleeping function called from invalid context at
../arch/x86/mm/fault.c:1187in_atomic(): 0,irqs_disabled(): 0, pid: 4706, name: bash Preemption disabled at:[<ffffffff81484339>]printk+0x48/0x4aTo fix this, we release the RCU read lock before we crash.
Tested this patch on linux 3.18 by booting off one of our boards.
Fixes 984d74a72076a1 ("sysrq: rcu-ify __handle_sysrq")Signed-off-by Ani Sinha <[email protected]>Reviewed-by Rik van Riel <[email protected]>Signed-off-by Paul E. McKenney <[email protected]>drivers/tty/sysrq.c +6
-0@@ -133,6 +133,12 @@ static void sysrq_handle_crash(int key) {char *killer =NULL ; +/* we need to release the RCU read lock here, +* otherwise we get an annoying +* 'BUG: sleeping function called from invalid context' +* complaint from the kernel before the panic. +*/ +rcu_read_unlock (); panic_on_oops =1 ;/* force panic */ wmb (); *killer =1 ;
Quoted 19 lines · Paul E. McKenney
Quoted 18 lines · Ani Sinha
Commit
984d74a72076a1("sysrq: rcu-ify__handle_sysrq") replacedspin_lock_irqsave()calls withrcu_read_lock()calls in sysrq. Sincercu_read_lock()does not disable preemption,faulthandler_disabled()in__do_page_fault()inx86/fault.creturns false. When the code later callsmight_sleep()in the pagefault handler, we get the following warning:BUG: sleeping function called from invalid context at ../arch/x86/mm/fault.c:1187
in_atomic(): 0,irqs_disabled(): 0, pid: 4706, name: bash Preemption disabled at:[<ffffffff81484339>]printk+0x48/0x4aTo fix this, we release the RCU read lock before we crash.
Tested this patch on linux 3.18 by booting off one of our boards.
Fixes 984d74a72076a1 ("sysrq: rcu-ify __handle_sysrq")
Quoted 9 lines
Quoted 1 line
Signed-off-by Ani Sinha <[email protected]>I queued this with Rik's Signed-off-by, and fixed some checkpatch.pl errors. Please run
scripts/checkpatch.plon your patches in the future.Please see below for the result.
Rik, did you test this as well? If so, may I also have your Tested-by?
I applied this patch on linux 4.4-rc5 and booted off a fedora core 22 vmware VM with it (sorry I don't have a real box around). Seems to be working fine. Also booted off a fc14 vm box with it and it also seems fine. Here's the kernel crash dump from fc22 box :
Fedora release 22 (Twenty Two)
Kernel 4.4.0-rc5-2847908.AroraKernelbleeding.5.fc18.x86_64 on an x86_64 (ttyS1)
localhost login: [ 50.070804] [drm:vmw_cmdbuf_work_func [vmwgfx]] *ERROR* Command buffer error.
[ 81.823507] sysrq: SysRq : Trigger a crash [ 81.825099] BUG: unable to handle kernel NULL pointer dereference at (null) [ 81.827357] IP: [<ffffffff8131c217>] sysrq_handle_crash+0x11/0x1b [ 81.828986] PGD 0 [ 81.829542] Oops: 0002 [#1] SMP [ 81.830426] Modules linked in: rfcomm xt_CHECKSUM ipt_MASQUERADE nf_nat_masquerade_ipv4 tun nf_conntrack_netbios_ns nf_conntrack_broadcast xt_tcpudp ip6t_REJECT fuse ipt_REJECT xt_conntrack ebtable_nat ebtable_broute bridge stp llc ebtable_filter ebtables ip6table_nat nf_conntrack_ipv6 nf_defrag_ipv6 nf_nat_ipv6 ip6table_mangle ip6table_security ip6table_raw ip6table_filter ip6_tables iptable_nat nf_conntrack_ipv4 nf_defrag_ipv4 nf_nat_ipv4 nf_nat nf_conntrack iptable_mangle iptable_security iptable_raw iptable_filter snd_seq_midi snd_seq_midi_event bnep coretemp hwmon crc32c_intel ppdev aesni_intel aes_x86_64 glue_helper lrw gf128mul ablk_helper cryptd snd_ens1371 snd_ac97_codec ac97_bus snd_seq pcspkr snd_pcm serio_raw snd_timer snd_rawmidi snd_seq_device snd soundcore gameport i2c_piix4 shpchp battery parport_pc parport btusb btrtl btbcm btintel ac bluetooth acpi_cpufreq tpm_tis tpm nfsd auth_rpcgss oid_registry nfs_acl lockd grace sunrpc ip_tables x_tables uhci_hcd vmwgfx e1000 ehci_pci ehci_hcd mptspi scsi_transport_spi mptscsih mptbase drm_kms_helper syscopyarea sysfillrect sysimgblt fb_sys_fops ttm drm i2c_core sr_mod cdrom dm_mirror dm_region_hash dm_log dm_mod autofs4 [ 81.854446] CPU: 0 PID: 2296 Comm: bash Not tainted 4.4.0-rc5-2847908.AroraKernelbleeding.5.fc18.x86_64 #1 [ 81.856338] Hardware name: VMware, Inc. VMware Virtual Platform/440BX Desktop Reference Platform, BIOS 6.00 05/20/2014 [ 81.858431] task: ffff88001c7d9a00 ti: ffff880004d70000 task.ti: ffff880004d70000 [ 81.859903] RIP: 0010:[<ffffffff8131c217>] [<ffffffff8131c217>] sysrq_handle_crash+0x11/0x1b [ 81.861609] RSP: 0018:ffff880004d73d98 EFLAGS: 00010296 [ 81.862672] RAX: 000000000000000f RBX: ffffffff81a774a0 RCX: 0000000000000000 [ 81.864078] RDX: ffff88003f60f101 RSI: ffff88003f60cae8 RDI: 0000000000000063 [ 81.865482] RBP: ffff880004d73d98 R08: 000000000000000f R09: 0000000000000000 [ 81.866889] R10: 0000000000000000 R11: 000000000000000f R12: 0000000000000004 [ 81.868294] R13: 0000000000000063 R14: ffff880004d73f00 R15: 0000000000000000 [ 81.869702] FS: 00007fd8e0c69700(0000) GS:ffff88003f600000(0000) knlGS:0000000000000000 [ 81.871293] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 81.872437] CR2: 0000000000000000 CR3: 000000001199f000 CR4: 00000000001406f0 [ 81.873877] Stack: [ 81.874297] ffff880004d73dc8 ffffffff8131c835 0000000000000002 ffff880004cb5200 [ 81.875876] 00007fd8e0c77000 ffff880004d73f20 ffff880004d73de8 ffffffff8131c8e7 [ 81.877455] fffffffffffffff2 ffff88003abff500 ffff880004d73e28 ffffffff81181949 [ 81.879032] Call Trace: [ 81.879537] [<ffffffff8131c835>] __handle_sysrq+0x8f/0x11c [ 81.880645] [<ffffffff8131c8e7>] write_sysrq_trigger+0x25/0x36 [ 81.881829] [<ffffffff81181949>] proc_reg_write+0x54/0x76 [ 81.883007] [<ffffffff81131ce4>] __vfs_write+0x23/0xa2 [ 81.884050] [<ffffffff81200200>] ? security_file_permission+0x37/0x40 [ 81.885344] [<ffffffff81132338>] ? rw_verify_area+0x6b/0xcb [ 81.886469] [<ffffffff81133ccf>] ? __sb_start_write+0x22/0x42 [ 81.887628] [<ffffffff81132abe>] vfs_write+0x86/0xdc [ 81.888634] [<ffffffff81132ca7>] SyS_write+0x4d/0x7f [ 81.889643] [<ffffffff8150406e>] entry_SYSCALL_64_fastpath+0x12/0x71 [ 81.890919] Code: 7a 00 00 55 b8 01 00 00 00 48 89 e5 75 07 0f b6 05 73 53 7a 00 83 e0 01 5d c3 55 c7 05 e7 23 92 00 01 00 00 00 48 89 e5 0f ae f8 <c6> 04 25 00 00 00 00 01 5d c3 55 48 89 e5 53 48 89 f3 48 8d 55 [ 81.896661] RIP [<ffffffff8131c217>] sysrq_handle_crash+0x11/0x1b [ 81.897915] RSP <ffff880004d73d98> [ 81.898618] CR2: 0000000000000000 [ 81.899346] ---[ end trace 7f48df1773107036 ]--- [ 81.900306] Kernel panic - not syncing: Fatal exception [ 81.901378] Kernel Offset: disabled [ 81.902109] ---[ end Kernel panic - not syncing: Fatal exception