* [PATCH] CacheFiles: Fix error handling in cachefiles_determine_cache_security()
@ 2010-05-12 14:34 David Howells
2010-05-12 19:36 ` Andrew Morton
2010-05-13 9:51 ` David Howells
0 siblings, 2 replies; 3+ messages in thread
From: David Howells @ 2010-05-12 14:34 UTC (permalink / raw)
To: torvalds, akpm; +Cc: dhowells, linux-cachefs, linux-kernel
cachefiles_determine_cache_security() is expected to return with a security
override in place. However, if set_create_files_as() fails, we fail to do
this. In this case, we should just reinstate the security override that was
set by the caller.
Furthermore, if set_create_files_as() fails, we should dispose of the new
credentials we were in the process of creating.
Signed-off-by: David Howells <dhowells@redhat.com>
---
fs/cachefiles/security.c | 4 ++++
1 files changed, 4 insertions(+), 0 deletions(-)
diff --git a/fs/cachefiles/security.c b/fs/cachefiles/security.c
index b5808cd..039b501 100644
--- a/fs/cachefiles/security.c
+++ b/fs/cachefiles/security.c
@@ -77,6 +77,8 @@ static int cachefiles_check_cache_dir(struct cachefiles_cache *cache,
/*
* check the security details of the on-disk cache
* - must be called with security override in force
+ * - must return with a security override in force - even in the case of an
+ * error
*/
int cachefiles_determine_cache_security(struct cachefiles_cache *cache,
struct dentry *root,
@@ -99,6 +101,8 @@ int cachefiles_determine_cache_security(struct cachefiles_cache *cache,
* which create files */
ret = set_create_files_as(new, root->d_inode);
if (ret < 0) {
+ abort_creds(new);
+ cachefiles_begin_secure(cache, _saved_cred);
_leave(" = %d [cfa]", ret);
return ret;
}
^ permalink raw reply related [flat|nested] 3+ messages in thread* Re: [PATCH] CacheFiles: Fix error handling in cachefiles_determine_cache_security()
2010-05-12 14:34 [PATCH] CacheFiles: Fix error handling in cachefiles_determine_cache_security() David Howells
@ 2010-05-12 19:36 ` Andrew Morton
2010-05-13 9:51 ` David Howells
1 sibling, 0 replies; 3+ messages in thread
From: Andrew Morton @ 2010-05-12 19:36 UTC (permalink / raw)
To: David Howells; +Cc: torvalds, linux-cachefs, linux-kernel
On Wed, 12 May 2010 15:34:03 +0100
David Howells <dhowells@redhat.com> wrote:
> cachefiles_determine_cache_security() is expected to return with a security
> override in place. However, if set_create_files_as() fails, we fail to do
> this. In this case, we should just reinstate the security override that was
> set by the caller.
>
> Furthermore, if set_create_files_as() fails, we should dispose of the new
> credentials we were in the process of creating.
>
> Signed-off-by: David Howells <dhowells@redhat.com>
> ---
>
> fs/cachefiles/security.c | 4 ++++
> 1 files changed, 4 insertions(+), 0 deletions(-)
>
> diff --git a/fs/cachefiles/security.c b/fs/cachefiles/security.c
> index b5808cd..039b501 100644
> --- a/fs/cachefiles/security.c
> +++ b/fs/cachefiles/security.c
> @@ -77,6 +77,8 @@ static int cachefiles_check_cache_dir(struct cachefiles_cache *cache,
> /*
> * check the security details of the on-disk cache
> * - must be called with security override in force
> + * - must return with a security override in force - even in the case of an
> + * error
> */
> int cachefiles_determine_cache_security(struct cachefiles_cache *cache,
> struct dentry *root,
> @@ -99,6 +101,8 @@ int cachefiles_determine_cache_security(struct cachefiles_cache *cache,
> * which create files */
> ret = set_create_files_as(new, root->d_inode);
> if (ret < 0) {
> + abort_creds(new);
> + cachefiles_begin_secure(cache, _saved_cred);
> _leave(" = %d [cfa]", ret);
> return ret;
> }
The changelog makes it hard for civilians to work out what the
user-visible effect of this bug was.
Should this fix be backported into -stable?
^ permalink raw reply [flat|nested] 3+ messages in thread* Re: [PATCH] CacheFiles: Fix error handling in cachefiles_determine_cache_security()
2010-05-12 14:34 [PATCH] CacheFiles: Fix error handling in cachefiles_determine_cache_security() David Howells
2010-05-12 19:36 ` Andrew Morton
@ 2010-05-13 9:51 ` David Howells
1 sibling, 0 replies; 3+ messages in thread
From: David Howells @ 2010-05-13 9:51 UTC (permalink / raw)
To: Andrew Morton; +Cc: dhowells, torvalds, linux-cachefs, linux-kernel
Andrew Morton <akpm@linux-foundation.org> wrote:
> The changelog makes it hard for civilians to work out what the
> user-visible effect of this bug was.
Sorry, I should've included the below oops report in the patch description.
Sadly, I think it's too late to fix that now as Linus has taken the patch.
The problem can only occur when the cache is being started up and if an LSM
(such as SELinux) is being used that can reject set_create_files_as(). It can
only be incurred by root, as only root can start up a cache.
Yes, it should got to stable too.
David
---
CacheFiles: Failed to register: -13
------------[ cut here ]------------
kernel BUG at kernel/cred.c:171!
invalid opcode: 0000 [#1] SMP
last sysfs file: /sys/block/sda/sda6/start
CPU 1
Modules linked in: cachefiles nfs fscache auth_rpcgss nfs_acl lockd sunrpc
Pid: 0, comm: swapper Not tainted 2.6.34-rc7-cachefs #74 DG965RY/
RIP: 0010:[<ffffffff81049bdf>] [<ffffffff81049bdf>] __put_cred+0x1a/0x65
RSP: 0018:ffff880002103e10 EFLAGS: 00010202
RAX: 0000000000000001 RBX: ffff88003dd94d80 RCX: 0000000000000000
RDX: ffff88003dc7d6e0 RSI: 0000000000000000 RDI: ffff88003dd94d80
RBP: ffff880002103e10 R08: 0000000000000050 R09: 09f911029d74e35b
R10: ffffffff8107ba45 R11: ffff88000210e640 R12: ffff880030670040
R13: ffff88000210e640 R14: ffff880030670f50 R15: ffff88003dc93d58
FS: 0000000000000000(0000) GS:ffff880002100000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: 00007f78e0e42098 CR3: 000000000161c000 CR4: 00000000000006e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Process swapper (pid: 0, threadinfo ffff88003dcb0000, task ffff88003dc84040)
Stack:
ffff880002103e30 ffffffff8104a78e ffff880030670040 ffffffff816326c0
<0> ffff880002103e50 ffffffff8102f87a ffff88000210e640 ffff880030670040
<0> ffff880002103e70 ffffffff810319d4 ffff880002103e70 ffff88000210e610
Call Trace:
<IRQ>
[<ffffffff8104a78e>] exit_creds+0x7f/0x15a
[<ffffffff8102f87a>] __put_task_struct+0x65/0x94
[<ffffffff810319d4>] delayed_put_task_struct+0x4e/0x52
[<ffffffff8107bb41>] __rcu_process_callbacks+0x191/0x2a4
[<ffffffff8107bc7a>] rcu_process_callbacks+0x26/0x4c
[<ffffffff810355e6>] __do_softirq+0xa2/0x140
[<ffffffff81002e0c>] call_softirq+0x1c/0x28
[<ffffffff81004b54>] do_softirq+0x38/0x80
[<ffffffff81034f82>] irq_exit+0x45/0x47
[<ffffffff81017833>] smp_apic_timer_interrupt+0x88/0x96
[<ffffffff810028d3>] apic_timer_interrupt+0x13/0x20
<EOI>
[<ffffffff810489e1>] ? __atomic_notifier_call_chain+0x0/0x86
[<ffffffff810096bf>] ? mwait_idle+0x6e/0x78
[<ffffffff810096b6>] ? mwait_idle+0x65/0x78
[<ffffffff810011dc>] cpu_idle+0x4d/0x83
[<ffffffff8138f62c>] start_secondary+0x1bd/0x1c1
Code: da c0 15 00 4c 89 e7 e8 bd 15 06 00 5b 41 5c c9 c3 55 8b 47 04 48 89 e5 8b 07 8b 07 85 c0 74 04 0f 0b eb fe 8b 47 04 85 c0 74 04 <0f> 0b eb fe c7 47 10 44 61 65 44 48 8b 45 08 48 89 47 08 65 48
RIP [<ffffffff81049bdf>] __put_cred+0x1a/0x65
RSP <ffff880002103e10>
---[ end trace f1e0ebe358730a1c ]---
Kernel panic - not syncing: Fatal exception in interrupt
Pid: 0, comm: swapper Tainted: G D 2.6.34-rc7-cachefs #74
Call Trace:
<IRQ> [<ffffffff81392977>] panic+0x73/0xeb
[<ffffffff81005c7c>] ? oops_end+0x42/0x8e
[<ffffffff81005cbb>] oops_end+0x81/0x8e
[<ffffffff81005e9f>] die+0x55/0x5e
[<ffffffff81003574>] do_trap+0x11c/0x12b
[<ffffffff810038fd>] ? do_invalid_op+0x72/0x9f
[<ffffffff81003921>] do_invalid_op+0x96/0x9f
[<ffffffff81049bdf>] ? __put_cred+0x1a/0x65
[<ffffffff81395418>] ? trace_hardirqs_off_thunk+0x3a/0x3c
[<ffffffff8107ba45>] ? __rcu_process_callbacks+0x95/0x2a4
[<ffffffff81395eec>] ? irq_return+0x0/0x2
[<ffffffff81002b95>] invalid_op+0x15/0x20
[<ffffffff8107ba45>] ? __rcu_process_callbacks+0x95/0x2a4
[<ffffffff81049bdf>] ? __put_cred+0x1a/0x65
[<ffffffff8104a78e>] exit_creds+0x7f/0x15a
[<ffffffff8102f87a>] __put_task_struct+0x65/0x94
[<ffffffff810319d4>] delayed_put_task_struct+0x4e/0x52
[<ffffffff8107bb41>] __rcu_process_callbacks+0x191/0x2a4
[<ffffffff8107bc7a>] rcu_process_callbacks+0x26/0x4c
[<ffffffff810355e6>] __do_softirq+0xa2/0x140
[<ffffffff81002e0c>] call_softirq+0x1c/0x28
[<ffffffff81004b54>] do_softirq+0x38/0x80
[<ffffffff81034f82>] irq_exit+0x45/0x47
[<ffffffff81017833>] smp_apic_timer_interrupt+0x88/0x96
[<ffffffff810028d3>] apic_timer_interrupt+0x13/0x20
<EOI> [<ffffffff810489e1>] ? __atomic_notifier_call_chain+0x0/0x86
[<ffffffff810096bf>] ? mwait_idle+0x6e/0x78
[<ffffffff810096b6>] ? mwait_idle+0x65/0x78
[<ffffffff810011dc>] cpu_idle+0x4d/0x83
[<ffffffff8138f62c>] start_secondary+0x1bd/0x1c1
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2010-05-13 9:51 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2010-05-12 14:34 [PATCH] CacheFiles: Fix error handling in cachefiles_determine_cache_security() David Howells
2010-05-12 19:36 ` Andrew Morton
2010-05-13 9:51 ` David Howells
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).