Date:	Sat, 20 Nov 2010 23:28:05 +0100 (CET)
From:	Jesper Juhl <jj@...osbits.net>
To:	Justin Mattock <justinmattock@...il.com>
cc:	Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: general protection fault: 0000 [#1] SMP

On Sat, 20 Nov 2010, Justin Mattock wrote:

> Ive seen this before, but could not reproduce for a bisect.. basically
> what I remember doing
> was building webkit(let sit and compile) passed out, woke up at 5AM
> closed the lid on the machine,few hrs later
> woke up, went for a run, came back opened the lid and this:
> 
> [43925.668053] general protection fault: 0000 [#1] SMP
> [43925.668059] last sysfs file: /sys/devices/platform/applesmc.768/light
> [43925.668061] CPU 0
> [43925.668063] Modules linked in: firewire_sbp2 radeon sco bnep ttm
> drm_kms_helper drm ipt_LOG iptable_nat nf_nat xt_state
> nf_conntrack_ftp nf_conntrack_ipv4 nf_conntrack nf_defrag_ipv4
> iptable_filter ip_tables x_tables ath9k ath9k_common video ath9k_hw
> sky2 firewire_ohci battery ac ath evdev joydev button firewire_core
> i2c_i801 kvm_intel aes_x86_64 lzo zlib ipcomp xfrm_ipcomp crypto_null
> sha256_generic cbc des_generic cast5 blowfish serpent camellia
> twofish_generic twofish_x86_64 twofish_common ctr ah4 esp4 authenc
> uhci_hcd ehci_hcd hci_uart rfcomm btusb hidp l2cap bluetooth coretemp
> acpi_cpufreq processor mperf appletouch applesmc uvcvideo
> [43925.668120]
> [43925.668123] Pid: 27262, comm: make Not tainted
> 2.6.37-rc2-00037-g7957f0a-dirty #6 Mac-F42187C8/MacBookPro2,2
> [43925.668126] RIP: 0010:[<ffffffff811bf10a>]  [<ffffffff811bf10a>]
> inode_has_perm+0x53/0x6a
> [43925.668135] RSP: 0018:ffff88003c5a5bc8  EFLAGS: 00010282
> [43925.668137] RAX: ffff88003826a208 RBX: ffff88000008ed80 RCX: ffff88003c5a5c68
> [43925.668140] RDX: 0000000000000002 RSI: ffff88000008ed80 RDI: ffff88002feacc00
> [43925.668142] RBP: ffff88003c5a5c58 R08: ffff88003c5a5c68 R09: 00000000000000d5
> [43925.668145] R10: 050366048b660e04 R11: 0000000000000000 R12: 0000000000000024
> [43925.668147] R13: 00000000ffffffd8 R14: 0000000000000000 R15: 0000000000000000
> [43925.668150] FS:  00007f4f786b3700(0000) GS:ffff88003ee00000(0000)
> knlGS:0000000000000000
> [43925.668153] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> [43925.668155] CR2: 00007f4f78637000 CR3: 00000000383ac000 CR4: 00000000000006e0
> [43925.668158] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
> [43925.668161] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
> [43925.668163] Process make (pid: 27262, threadinfo ffff88003c5a4000,
> task ffff880001afb410)
> [43925.668165] Stack:
> [43925.668167]  ffff880038a98060 0000000000000000 ffff88003c5a5c48
> ffffffff81182b7c
> [43925.668172]  ffff88003cab2688 ffff880024da9990 ffff88003caa18d8
> ffff880038a98060
> [43925.668177]  ffff880024da98b0 ffffea0000a54940 ffff88003c5a5c78
> ffff88003d402500
> [43925.668182] Call Trace:
> [43925.668189]  [<ffffffff81182b7c>] ? jbd2_journal_stop+0x21e/0x230
> [43925.668193]  [<ffffffff811be4bb>] ? selinux_cred_free+0xb/0x27
> [43925.668196]  [<ffffffff811be441>] ? selinux_file_alloc_security+0x4a/0xb9
> [43925.668201]  [<ffffffff810f4226>] ? check_object+0x13b/0x1eb
> [43925.668205]  [<ffffffff811bf853>] selinux_inode_permission+0xd2/0xd4
> [43925.668211]  [<ffffffff811bbf9c>] security_inode_permission+0x1c/0x1e
> [43925.668215]  [<ffffffff81101ab2>] inode_permission+0x87/0x93
> [43925.668218]  [<ffffffff81102e86>] may_open+0x9e/0x11e
> [43925.668221]  [<ffffffff8110373e>] do_last+0x542/0x6fa
> [43925.668225]  [<ffffffff811056ec>] do_filp_open+0x1f3/0x646
> [43925.668228]  [<ffffffff810f4226>] ? check_object+0x13b/0x1eb
> [43925.668232]  [<ffffffff81103958>] ? getname+0x2c/0x1be
> [43925.668236]  [<ffffffff8110eca8>] ? alloc_fd+0x111/0x123
> [43925.668240]  [<ffffffff810f7a84>] do_sys_open+0x5b/0xf8
> [43925.668243]  [<ffffffff810f7b4a>] sys_open+0x1b/0x1d
> [43925.668248]  [<ffffffff8102b542>] system_call_fastpath+0x16/0x1b
> [43925.668250] Code: 02 00 00 44 8b 48 04 48 85 c9 75 1f 4c 8d 85 70
> ff ff ff b9 22 00 00 00 4c 89 c7 44 89 d8 f3 ab c6 85 70 ff ff ff 01
> 48 89 75 90 <41> 0f b7 42 20 89 d1 41 8b 72 1c 89 c2 44 89 cf e8 99 e7
> ff ff
> [43925.668288] RIP  [<ffffffff811bf10a>] inode_has_perm+0x53/0x6a
> [43925.668291]  RSP <ffff88003c5a5bc8>
> [43925.668295] ---[ end trace 75bdddc506717838 ]---
> [43934.866252] general protection fault: 0000 [#2] SMP
> [43934.866257] last sysfs file: /sys/devices/platform/applesmc.768/light
> [43934.866260] CPU 0
> [43934.866261] Modules linked in: firewire_sbp2 radeon sco bnep ttm
> drm_kms_helper drm ipt_LOG iptable_nat nf_nat xt_state
> nf_conntrack_ftp nf_conntrack_ipv4 nf_conntrack nf_defrag_ipv4
> iptable_filter ip_tables x_tables ath9k ath9k_common video ath9k_hw
> sky2 firewire_ohci battery ac ath evdev joydev button firewire_core
> i2c_i801 kvm_intel aes_x86_64 lzo zlib ipcomp xfrm_ipcomp crypto_null
> sha256_generic cbc des_generic cast5 blowfish serpent camellia
> twofish_generic twofish_x86_64 twofish_common ctr ah4 esp4 authenc
> uhci_hcd ehci_hcd hci_uart rfcomm btusb hidp l2cap bluetooth coretemp
> acpi_cpufreq processor mperf appletouch applesmc uvcvideo
> [43934.866318]
> [43934.866321] Pid: 27283, comm: make Tainted: G      D
> 2.6.37-rc2-00037-g7957f0a-dirty #6 Mac-F42187C8/MacBookPro2,2
> [43934.866324] RIP: 0010:[<ffffffff811bf10a>]  [<ffffffff811bf10a>]
> inode_has_perm+0x53/0x6a
> [43934.866334] RSP: 0018:ffff88003c5a5bc8  EFLAGS: 00010282
> [43934.866336] RAX: ffff88003807a958 RBX: ffff88000008ed80 RCX: ffff88003c5a5c68
> [43934.866339] RDX: 0000000000000002 RSI: ffff88000008ed80 RDI: ffff880034b01700
> [43934.866341] RBP: ffff88003c5a5c58 R08: ffff88003c5a5c68 R09: 00000000000000d5
> [43934.866343] R10: 050366048b660e04 R11: 0000000000000000 R12: 0000000000000024
> [43934.866346] R13: 00000000ffffffd8 R14: 0000000000000000 R15: 0000000000000000
> [43934.866349] FS:  00007fdf0a661700(0000) GS:ffff88003ee00000(0000)
> knlGS:0000000000000000
> [43934.866352] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> [43934.866354] CR2: 00007fdf0a5e5000 CR3: 0000000029800000 CR4: 00000000000006e0
> [43934.866357] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
> [43934.866359] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
> [43934.866362] Process make (pid: 27283, threadinfo ffff88003c5a4000,
> task ffff880001afb410)
> [43934.866364] Stack:
> [43934.866366]  ffff88002f398a50 ffff880024da9990 000000003c5a5c78
> ffffffff81810be8
> [43934.866371]  0020000000000001 0000000000000001 0000000000001000
> ffff880037bc0a00
> [43934.866375]  0000000000001000 ffffea0000a54940 ffff88003c5a5d18
> ffff88003d402500
> [43934.866380] Call Trace:
> [43934.866385]  [<ffffffff811be4bb>] ? selinux_cred_free+0xb/0x27
> [43934.866389]  [<ffffffff811be441>] ? selinux_file_alloc_security+0x4a/0xb9
> [43934.866395]  [<ffffffff810f4226>] ? check_object+0x13b/0x1eb
> [43934.866398]  [<ffffffff811bf853>] selinux_inode_permission+0xd2/0xd4
> [43934.866404]  [<ffffffff811bbf9c>] security_inode_permission+0x1c/0x1e
> [43934.866409]  [<ffffffff81101ab2>] inode_permission+0x87/0x93
> [43934.866412]  [<ffffffff81102e86>] may_open+0x9e/0x11e
> [43934.866415]  [<ffffffff8110373e>] do_last+0x542/0x6fa
> [43934.866419]  [<ffffffff811056ec>] do_filp_open+0x1f3/0x646
> [43934.866422]  [<ffffffff810f4226>] ? check_object+0x13b/0x1eb
> [43934.866426]  [<ffffffff81103958>] ? getname+0x2c/0x1be
> [43934.866430]  [<ffffffff8110eca8>] ? alloc_fd+0x111/0x123
> [43934.866433]  [<ffffffff810f7a84>] do_sys_open+0x5b/0xf8
> [43934.866437]  [<ffffffff810f7b4a>] sys_open+0x1b/0x1d
> [43934.866441]  [<ffffffff8102b542>] system_call_fastpath+0x16/0x1b
> [43934.866443] Code: 02 00 00 44 8b 48 04 48 85 c9 75 1f 4c 8d 85 70
> ff ff ff b9 22 00 00 00 4c 89 c7 44 89 d8 f3 ab c6 85 70 ff ff ff 01
> 48 89 75 90 <41> 0f b7 42 20 89 d1 41 8b 72 1c 89 c2 44 89 cf e8 99 e7
> ff ff
> [43934.866481] RIP  [<ffffffff811bf10a>] inode_has_perm+0x53/0x6a
> [43934.866484]  RSP <ffff88003c5a5bc8>
> [43934.866488] ---[ end trace 75bdddc506717839 ]---
> 
[...]

Hmm, ok, I have no idea about the root cause of this problem, but I did 
notice one thing about selinux_cred_free() that's different than most 
other freeing functions in the kernel. It does not accept a NULL value.
Most other freeing functions will just return if passed NULL, but 
selinux_cred_free() will crash.
I wonder if it would make sense to add a NULL 'short circuit' to that 
function? If so, please pick up the patch below.


Signed-off-by: Jesper Juhl <jj@...osbits.net>
---
 hooks.c |    6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c
index 65fa8bf..d088532 100644
--- a/security/selinux/hooks.c
+++ b/security/selinux/hooks.c
@@ -3193,11 +3193,11 @@ static int selinux_cred_alloc_blank(struct cred *cred, gfp_t gfp)
  */
 static void selinux_cred_free(struct cred *cred)
 {
-	struct task_security_struct *tsec = cred->security;
-
+	if (!cred)
+		return;
 	BUG_ON((unsigned long) cred->security < PAGE_SIZE);
 	cred->security = (void *) 0x7UL;
-	kfree(tsec);
+	kfree(cred->security);
 }
 
 /*


-- 
Jesper Juhl <jj@...osbits.net>            http://www.chaosbits.net/
Don't top-post http://www.catb.org/~esr/jargon/html/T/top-post.html
Plain text mails only, please.

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Powered by blists - more mailing lists