nfs: handle failure of nfs_get_lock_context in unlock path
Subsystems (2)
PATCH v2 1/1 nfs: handle failure of nfs_get_lock_context in unlock path
When memory is insufficient, the allocation of nfs_lock_context in nfs_get_lock_context() fails and returns -ENOMEM. If we mistakenly treat an nfs4_unlockdata structure (whose l_ctx member has been set to -ENOMEM) as valid and proceed to execute rpc_run_task(), this will trigger a NULL pointer dereference in nfs4_locku_prepare. For example:
BUG: kernel NULL pointer dereference, address: 000000000000000c PGD 0 P4D 0 Oops: Oops: 0000 [#1] SMP PTI CPU: 15 UID: 0 PID: 12 Comm: kworker/u64:0 Not tainted 6.15.0-rc2-dirty #60 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.16.3-2.fc40 Workqueue: rpciod rpc_async_schedule RIP: 0010:nfs4_locku_prepare+0x35/0xc2 Code: 89 f2 48 89 fd 48 c7 c7 68 69 ef b5 53 48 8b 8e 90 00 00 00 48 89 f3 RSP: 0018:ffffbbafc006bdb8 EFLAGS: 00010246 RAX: 000000000000004b RBX: ffff9b964fc1fa00 RCX: 0000000000000000 RDX: 0000000000000000 RSI: fffffffffffffff4 RDI: ffff9ba53fddbf40 RBP: ffff9ba539934000 R08: 0000000000000000 R09: ffffbbafc006bc38 R10: ffffffffb6b689c8 R11: 0000000000000003 R12: ffff9ba539934030 R13: 0000000000000001 R14: 0000000004248060 R15: ffffffffb56d1c30 FS: 0000000000000000(0000) GS:ffff9ba5881f0000(0000) knlGS:00000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000000000000000c CR3: 000000093f244000 CR4: 00000000000006f0 Call Trace: <TASK> __rpc_execute+0xbc/0x480 rpc_async_schedule+0x2f/0x40 process_one_work+0x232/0x5d0 worker_thread+0x1da/0x3d0 ? __pfx_worker_thread+0x10/0x10 kthread+0x10d/0x240 ? __pfx_kthread+0x10/0x10 ret_from_fork+0x34/0x50 ? __pfx_kthread+0x10/0x10 ret_from_fork_asm+0x1a/0x30 </TASK> Modules linked in: CR2: 000000000000000c ---[ end trace 0000000000000000 ]---
Free the allocated nfs4_unlockdata when nfs_get_lock_context() fails and return NULL to terminate subsequent rpc_run_task, preventing NULL pointer dereference.
---
Changes in v2:
Add a comment explaining that error handling for ctx acquisition failure is unnecessary.
fs/nfs/nfs4proc.c | 13 ++++++++++++- 1 file changed, 12 insertions(+), 1 deletion(-)
fs/nfs/nfs4proc.c +12 -1
@@ -7074,18 +7074,29 @@ static struct nfs4_unlockdata *nfs4_alloc_unlockdata(struct file_lock *fl,struct nfs4_unlockdata *p;struct nfs4_state *state = lsp->ls_state;struct inode *inode = state->inode; +struct nfs_lock_context *l_ctx; p =kzalloc (sizeof (*p),GFP_KERNEL );if (p ==NULL )return NULL ; + l_ctx =nfs_get_lock_context (ctx); +if (!IS_ERR (l_ctx)) { + p->l_ctx = l_ctx; + }else { +kfree (p); +return NULL ; + } p->arg.fh =NFS_FH (inode); p->arg.fl = &p->fl; p->arg.seqid = seqid; p->res.seqid = seqid; p->lsp = lsp;/* Ensure we don't close file until we're done freeing locks! */ +/* +* Since the caller holds a reference to ctx, the refcount must be non-zero. +* Therefore, error handling for failed ctx acquisition is unnecessary here. +*/ p->ctx =get_nfs_open_context (ctx); - p->l_ctx =nfs_get_lock_context (ctx);locks_init_lock (&p->fl);locks_copy_lock (&p->fl, fl); p->server =NFS_SERVER (inode);
Signature · 1 line
2.31.1
Quoted 75 lines · Li Lingfeng
When memory is insufficient, the allocation of
nfs_lock_contextinnfs_get_lock_context()fails and returns-ENOMEM. If we mistakenly treat annfs4_unlockdatastructure (whosel_ctxmember has been set to-ENOMEM) as valid and proceed to executerpc_run_task(), this will trigger a NULL pointer dereference innfs4_locku_prepare. For example:BUG: kernel NULL pointer dereference, address: 000000000000000c PGD 0 P4D 0 Oops: Oops: 0000 [#1] SMP PTI CPU: 15 UID: 0 PID: 12 Comm: kworker/u64:0 Not tainted 6.15.0-rc2-dirty #60 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.16.3-2.fc40 Workqueue: rpciod rpc_async_schedule RIP: 0010:nfs4_locku_prepare+0x35/0xc2 Code: 89 f2 48 89 fd 48 c7 c7 68 69 ef b5 53 48 8b 8e 90 00 00 00 48 89 f3 RSP: 0018:ffffbbafc006bdb8 EFLAGS: 00010246 RAX: 000000000000004b RBX: ffff9b964fc1fa00 RCX: 0000000000000000 RDX: 0000000000000000 RSI: fffffffffffffff4 RDI: ffff9ba53fddbf40 RBP: ffff9ba539934000 R08: 0000000000000000 R09: ffffbbafc006bc38 R10: ffffffffb6b689c8 R11: 0000000000000003 R12: ffff9ba539934030 R13: 0000000000000001 R14: 0000000004248060 R15: ffffffffb56d1c30 FS: 0000000000000000(0000) GS:ffff9ba5881f0000(0000) knlGS:00000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000000000000000c CR3: 000000093f244000 CR4: 00000000000006f0 Call Trace: <TASK> __rpc_execute+0xbc/0x480 rpc_async_schedule+0x2f/0x40 process_one_work+0x232/0x5d0 worker_thread+0x1da/0x3d0 ? __pfx_worker_thread+0x10/0x10 kthread+0x10d/0x240 ? __pfx_kthread+0x10/0x10 ret_from_fork+0x34/0x50 ? __pfx_kthread+0x10/0x10 ret_from_fork_asm+0x1a/0x30 </TASK> Modules linked in: CR2: 000000000000000c ---[ end trace 0000000000000000 ]---Free the allocated
nfs4_unlockdatawhennfs_get_lock_context()fails and return NULL to terminate subsequentrpc_run_task, preventing NULL pointer dereference.Fixes f30cb757f680 ("NFS: Always wait for I/O completion before unlock")Signed-off-by Li Lingfeng <[email protected]>---
Changes in v2:
Add a comment explaining that error handling for ctx acquisition failure is unnecessary.fs/nfs/nfs4proc.c | 13 ++++++++++++- 1 file changed, 12 insertions(+), 1 deletion(-)fs/nfs/nfs4proc.c +8
-0@@ -7074,18 +7074,29 @@ static struct nfs4_unlockdata *nfs4_alloc_unlockdata(struct file_lock *fl,struct nfs4_unlockdata *p;struct nfs4_state *state = lsp->ls_state;struct inode *inode = state->inode; +struct nfs_lock_context *l_ctx; p =kzalloc (sizeof (*p),GFP_KERNEL );if (p ==NULL )return NULL ; + l_ctx =nfs_get_lock_context (ctx); +if (!IS_ERR (l_ctx)) { + p->l_ctx = l_ctx; + }else { +kfree (p); +return NULL ; + }
fs/nfs/nfs4proc.c static struct nfs4_unlockdata *nfs4_alloc_unlockdata(struct file_lock *fl, · line 7098
Quoted 10 lines
fs/nfs/nfs4proc.c +4
-0p->arg.fh =NFS_FH (inode); p->arg.fl = &p->fl; p->arg.seqid = seqid; p->res.seqid = seqid; p->lsp = lsp;/* Ensure we don't close file until we're done freeing locks! */ +/* +* Since the caller holds a reference to ctx, the refcount must be non-zero. +* Therefore, error handling for failed ctx acquisition is unnecessary here. +*/
I think the additional comment is unnecessary, but otherwise this looks correct.
Ben
Quoted 80 lines · Li Lingfeng
When memory is insufficient, the allocation of
nfs_lock_contextinnfs_get_lock_context()fails and returns-ENOMEM. If we mistakenly treat annfs4_unlockdatastructure (whosel_ctxmember has been set to-ENOMEM) as valid and proceed to executerpc_run_task(), this will trigger a NULL pointer dereference innfs4_locku_prepare. For example:BUG: kernel NULL pointer dereference, address: 000000000000000c PGD 0 P4D 0 Oops: Oops: 0000 [#1] SMP PTI CPU: 15 UID: 0 PID: 12 Comm: kworker/u64:0 Not tainted 6.15.0-rc2-dirty #60 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.16.3-2.fc40 Workqueue: rpciod rpc_async_schedule RIP: 0010:nfs4_locku_prepare+0x35/0xc2 Code: 89 f2 48 89 fd 48 c7 c7 68 69 ef b5 53 48 8b 8e 90 00 00 00 48 89 f3 RSP: 0018:ffffbbafc006bdb8 EFLAGS: 00010246 RAX: 000000000000004b RBX: ffff9b964fc1fa00 RCX: 0000000000000000 RDX: 0000000000000000 RSI: fffffffffffffff4 RDI: ffff9ba53fddbf40 RBP: ffff9ba539934000 R08: 0000000000000000 R09: ffffbbafc006bc38 R10: ffffffffb6b689c8 R11: 0000000000000003 R12: ffff9ba539934030 R13: 0000000000000001 R14: 0000000004248060 R15: ffffffffb56d1c30 FS: 0000000000000000(0000) GS:ffff9ba5881f0000(0000) knlGS:00000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000000000000000c CR3: 000000093f244000 CR4: 00000000000006f0 Call Trace: <TASK> __rpc_execute+0xbc/0x480 rpc_async_schedule+0x2f/0x40 process_one_work+0x232/0x5d0 worker_thread+0x1da/0x3d0 ? __pfx_worker_thread+0x10/0x10 kthread+0x10d/0x240 ? __pfx_kthread+0x10/0x10 ret_from_fork+0x34/0x50 ? __pfx_kthread+0x10/0x10 ret_from_fork_asm+0x1a/0x30 </TASK> Modules linked in: CR2: 000000000000000c ---[ end trace 0000000000000000 ]---Free the allocated
nfs4_unlockdatawhennfs_get_lock_context()fails and return NULL to terminate subsequentrpc_run_task, preventing NULL pointer dereference.Fixes f30cb757f680 ("NFS: Always wait for I/O completion before unlock")Signed-off-by Li Lingfeng <[email protected]>---
Changes in v2:
Add a comment explaining that error handling for ctx acquisition failure is unnecessary.fs/nfs/nfs4proc.c | 13 ++++++++++++- 1 file changed, 12 insertions(+), 1 deletion(-)fs/nfs/nfs4proc.c +8
-0@@ -7074,18 +7074,29 @@ static struct nfs4_unlockdata *nfs4_alloc_unlockdata(struct file_lock *fl,struct nfs4_unlockdata *p;struct nfs4_state *state = lsp->ls_state;struct inode *inode = state->inode; +struct nfs_lock_context *l_ctx; p =kzalloc (sizeof (*p),GFP_KERNEL );if (p ==NULL )return NULL ; + l_ctx =nfs_get_lock_context (ctx); +if (!IS_ERR (l_ctx)) { + p->l_ctx = l_ctx; + }else { +kfree (p); +return NULL ; + } p->arg.fh =NFS_FH (inode); p->arg.fl = &p->fl; p->arg.seqid = seqid; p->res.seqid = seqid; p->lsp = lsp;
fs/nfs/nfs4proc.c static struct nfs4_unlockdata *nfs4_alloc_unlockdata(struct file_lock *fl, · line 7102
Quoted 10 lines
fs/nfs/nfs4proc.c +4
-1/* Ensure we don't close file until we're done freeing locks! */ +/* +* Since the caller holds a reference to ctx, the refcount must be non-zero. +* Therefore, error handling for failed ctx acquisition is unnecessary here. +*/ p->ctx =get_nfs_open_context (ctx); - p->l_ctx =nfs_get_lock_context (ctx);locks_init_lock (&p->fl);locks_copy_lock (&p->fl, fl); p->server =NFS_SERVER (inode);
Hi Li,
Quoted 44 lines · Li Lingfeng
When memory is insufficient, the allocation of
nfs_lock_contextinnfs_get_lock_context()fails and returns-ENOMEM. If we mistakenly treat annfs4_unlockdatastructure (whosel_ctxmember has been set to-ENOMEM) as valid and proceed to executerpc_run_task(), this will trigger a NULL pointer dereference innfs4_locku_prepare. For example:BUG: kernel NULL pointer dereference, address: 000000000000000c PGD 0 P4D 0 Oops: Oops: 0000 [#1] SMP PTI CPU: 15 UID: 0 PID: 12 Comm: kworker/u64:0 Not tainted 6.15.0-rc2-dirty #60 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.16.3-2.fc40 Workqueue: rpciod rpc_async_schedule RIP: 0010:nfs4_locku_prepare+0x35/0xc2 Code: 89 f2 48 89 fd 48 c7 c7 68 69 ef b5 53 48 8b 8e 90 00 00 00 48 89 f3 RSP: 0018:ffffbbafc006bdb8 EFLAGS: 00010246 RAX: 000000000000004b RBX: ffff9b964fc1fa00 RCX: 0000000000000000 RDX: 0000000000000000 RSI: fffffffffffffff4 RDI: ffff9ba53fddbf40 RBP: ffff9ba539934000 R08: 0000000000000000 R09: ffffbbafc006bc38 R10: ffffffffb6b689c8 R11: 0000000000000003 R12: ffff9ba539934030 R13: 0000000000000001 R14: 0000000004248060 R15: ffffffffb56d1c30 FS: 0000000000000000(0000) GS:ffff9ba5881f0000(0000) knlGS:00000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000000000000000c CR3: 000000093f244000 CR4: 00000000000006f0 Call Trace: <TASK> __rpc_execute+0xbc/0x480 rpc_async_schedule+0x2f/0x40 process_one_work+0x232/0x5d0 worker_thread+0x1da/0x3d0 ? __pfx_worker_thread+0x10/0x10 kthread+0x10d/0x240 ? __pfx_kthread+0x10/0x10 ret_from_fork+0x34/0x50 ? __pfx_kthread+0x10/0x10 ret_from_fork_asm+0x1a/0x30 </TASK> Modules linked in: CR2: 000000000000000c ---[ end trace 0000000000000000 ]---Free the allocated
nfs4_unlockdatawhennfs_get_lock_context()fails and return NULL to terminate subsequentrpc_run_task, preventing NULL pointer dereference.
Quoted 7 lines
Fixes f30cb757f680 ("NFS: Always wait for I/O completion before unlock")Signed-off-by Li Lingfeng <[email protected]>---
Changes in v2:
Add a comment explaining that error handling for ctx acquisition failure is unnecessary.
Thanks for the patch!
Quoted 14 lines
fs/nfs/nfs4proc.c | 13 ++++++++++++- 1 file changed, 12 insertions(+), 1 deletion(-)fs/nfs/nfs4proc.c +1
-0@@ -7074,18 +7074,29 @@ static struct nfs4_unlockdata *nfs4_alloc_unlockdata(struct file_lock *fl,struct nfs4_unlockdata *p;struct nfs4_state *state = lsp->ls_state;struct inode *inode = state->inode; +struct nfs_lock_context *l_ctx;
fs/nfs/nfs4proc.c static struct nfs4_unlockdata *nfs4_alloc_unlockdata(struct file_lock *fl, · line 7088
Quoted 10 lines
fs/nfs/nfs4proc.c +7
-0p =kzalloc (sizeof (*p),GFP_KERNEL );if (p ==NULL )return NULL ; + l_ctx =nfs_get_lock_context (ctx); +if (!IS_ERR (l_ctx)) { + p->l_ctx = l_ctx; + }else { +kfree (p); +return NULL ; + }
I was thinking the code would look a little neater if this if / else block was changed to:
if (IS_ERR(l_ctx)) {
kfree(p);
return NULL;
}fs/nfs/nfs4proc.c static struct nfs4_unlockdata *nfs4_alloc_unlockdata(struct file_lock *fl, · line 7088
Quoted 12 lines
fs/nfs/nfs4proc.c +4
-1p->arg.fh =NFS_FH (inode); p->arg.fl = &p->fl; p->arg.seqid = seqid; p->res.seqid = seqid; p->lsp = lsp;/* Ensure we don't close file until we're done freeing locks! */ +/* +* Since the caller holds a reference to ctx, the refcount must be non-zero. +* Therefore, error handling for failed ctx acquisition is unnecessary here. +*/ p->ctx =get_nfs_open_context (ctx); - p->l_ctx =nfs_get_lock_context (ctx);
And then update this to set p->l_ctx = l_ctx.
What do you think?
Anna
Quoted 3 lines
fs/nfs/nfs4proc.c +0
-0locks_init_lock (&p->fl);locks_copy_lock (&p->fl, fl); p->server =NFS_SERVER (inode);
Hi Anna,
Thank you for reviewing the patch.
I agree that handling the error case first makes the code cleaner and reduces nesting. Your suggested change looks great to me. I'll send v3 soon.
Lingfeng
Quoted 108 lines · Anna Schumaker
Hi Li,
Quoted 44 lines · Li Lingfeng
When memory is insufficient, the allocation of
nfs_lock_contextinnfs_get_lock_context()fails and returns-ENOMEM. If we mistakenly treat annfs4_unlockdatastructure (whosel_ctxmember has been set to-ENOMEM) as valid and proceed to executerpc_run_task(), this will trigger a NULL pointer dereference innfs4_locku_prepare. For example:BUG: kernel NULL pointer dereference, address: 000000000000000c PGD 0 P4D 0 Oops: Oops: 0000 [#1] SMP PTI CPU: 15 UID: 0 PID: 12 Comm: kworker/u64:0 Not tainted 6.15.0-rc2-dirty #60 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.16.3-2.fc40 Workqueue: rpciod rpc_async_schedule RIP: 0010:nfs4_locku_prepare+0x35/0xc2 Code: 89 f2 48 89 fd 48 c7 c7 68 69 ef b5 53 48 8b 8e 90 00 00 00 48 89 f3 RSP: 0018:ffffbbafc006bdb8 EFLAGS: 00010246 RAX: 000000000000004b RBX: ffff9b964fc1fa00 RCX: 0000000000000000 RDX: 0000000000000000 RSI: fffffffffffffff4 RDI: ffff9ba53fddbf40 RBP: ffff9ba539934000 R08: 0000000000000000 R09: ffffbbafc006bc38 R10: ffffffffb6b689c8 R11: 0000000000000003 R12: ffff9ba539934030 R13: 0000000000000001 R14: 0000000004248060 R15: ffffffffb56d1c30 FS: 0000000000000000(0000) GS:ffff9ba5881f0000(0000) knlGS:00000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000000000000000c CR3: 000000093f244000 CR4: 00000000000006f0 Call Trace: <TASK> __rpc_execute+0xbc/0x480 rpc_async_schedule+0x2f/0x40 process_one_work+0x232/0x5d0 worker_thread+0x1da/0x3d0 ? __pfx_worker_thread+0x10/0x10 kthread+0x10d/0x240 ? __pfx_kthread+0x10/0x10 ret_from_fork+0x34/0x50 ? __pfx_kthread+0x10/0x10 ret_from_fork_asm+0x1a/0x30 </TASK> Modules linked in: CR2: 000000000000000c ---[ end trace 0000000000000000 ]---Free the allocated
nfs4_unlockdatawhennfs_get_lock_context()fails and return NULL to terminate subsequentrpc_run_task, preventing NULL pointer dereference.Quoted 7 lines
Fixes f30cb757f680 ("NFS: Always wait for I/O completion before unlock")Signed-off-by Li Lingfeng <[email protected]>---
Changes in v2:
Add a comment explaining that error handling for ctx acquisition failure is unnecessary.Thanks for the patch!
Quoted 13 lines
fs/nfs/nfs4proc.c | 13 ++++++++++++- 1 file changed, 12 insertions(+), 1 deletion(-)fs/nfs/nfs4proc.c +1
-0@@ -7074,18 +7074,29 @@ static struct nfs4_unlockdata *nfs4_alloc_unlockdata(struct file_lock *fl,struct nfs4_unlockdata *p;struct nfs4_state *state = lsp->ls_state;struct inode *inode = state->inode; +struct nfs_lock_context *l_ctx;fs/nfs/nfs4proc.c static struct nfs4_unlockdata *nfs4_alloc_unlockdata(struct file_lock *fl, · line 7088
Quoted 10 lines
fs/nfs/nfs4proc.c +7
-0p =kzalloc (sizeof (*p),GFP_KERNEL );if (p ==NULL )return NULL ; + l_ctx =nfs_get_lock_context (ctx); +if (!IS_ERR (l_ctx)) { + p->l_ctx = l_ctx; + }else { +kfree (p); +return NULL ; + }I was thinking the code would look a little neater if this if / else block was changed to:
if (IS_ERR(l_ctx)) { kfree(p); return NULL; }fs/nfs/nfs4proc.c static struct nfs4_unlockdata *nfs4_alloc_unlockdata(struct file_lock *fl, · line 7088
Quoted 12 lines
fs/nfs/nfs4proc.c +4
-1p->arg.fh =NFS_FH (inode); p->arg.fl = &p->fl; p->arg.seqid = seqid; p->res.seqid = seqid; p->lsp = lsp;/* Ensure we don't close file until we're done freeing locks! */ +/* +* Since the caller holds a reference to ctx, the refcount must be non-zero. +* Therefore, error handling for failed ctx acquisition is unnecessary here. +*/ p->ctx =get_nfs_open_context (ctx); - p->l_ctx =nfs_get_lock_context (ctx);And then update this to set
p->l_ctx=l_ctx.What do you think?
AnnaQuoted 3 lines
fs/nfs/nfs4proc.c +0
-0locks_init_lock (&p->fl);locks_copy_lock (&p->fl, fl); p->server =NFS_SERVER (inode);