ocfs2: validate group add input before caching
PATCH 1/1 ocfs2: validate group add input before caching
[BUG]OCFS2_IOC_GROUP_ADD can trigger a BUG_ON inocfs2_set_new_buffer_uptodate():
kernel BUG at fs/ocfs2/uptodate.c:509!
Oops: invalid opcode: 0000 [#1] SMP KASAN NOPTI RIP: 0010:ocfs2_set_new_buffer_uptodate+0x194/0x1e0 fs/ocfs2/uptodate.c:509 Code: ffffe88f 42b9fe4c 89e64889 dfe8b4df Call Trace: ocfs2_group_add+0x3f1/0x1510 fs/ocfs2/resize.c:507 ocfs2_ioctl+0x309/0x6e0 fs/ocfs2/ioctl.c:887 vfs_ioctl fs/ioctl.c:51 [inline] __do_sys_ioctl fs/ioctl.c:597 [inline] __se_sys_ioctl fs/ioctl.c:583 [inline] __x64_sys_ioctl+0x197/0x1e0 fs/ioctl.c:583 x64_sys_call+0x1144/0x26a0 arch/x86/include/generated/asm/syscalls_64.h:17 do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline] do_syscall_64+0x93/0xf80 arch/x86/entry/syscall_64.c:94 entry_SYSCALL_64_after_hwframe+0x76/0x7e RIP: 0033:0x7bbfb55a966d
[CAUSE]ocfs2_group_add() calls ocfs2_set_new_buffer_uptodate() on a user-controlled group block before ocfs2_verify_group_and_input() validates that block number. That helper is only valid for newly allocated metadata and asserts that the block is not already present in the chosen metadata cache. The code also uses INODE_CACHE(inode) even though the group descriptor belongs to main_bm_inode and later journal accesses use that cache context instead.
[FIX]
Validate the on-disk group descriptor before caching it, then add it to the metadata cache tracked by INODE_CACHE(main_bm_inode). Update the error path to remove the buffer from the same cache context so the group buffer lifetime stays consistent across validation, journaling, and cleanup.
--- fs/ocfs2/resize.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
fs/ocfs2/resize.c +3 -3
@@ -504,14 +504,14 @@ int ocfs2_group_add(struct inode *inode, struct ocfs2_new_group_input *input) goto out_unlock; } -ocfs2_set_new_buffer_uptodate (INODE_CACHE (inode), group_bh); - ret =ocfs2_verify_group_and_input (main_bm_inode, fe, input, group_bh);if (ret) {mlog_errno (ret); goto out_free_group_bh; } +ocfs2_set_new_buffer_uptodate (INODE_CACHE (main_bm_inode), group_bh); +trace_ocfs2_group_add ((unsigned long long )input->group, input->chain, input->clusters, input->frees);
@@ -575,7 +575,7 @@ int ocfs2_group_add(struct inode *inode, struct ocfs2_new_group_input *input) out_free_group_bh:if (ret <0 ) -ocfs2_remove_from_cache (INODE_CACHE (inode), group_bh); +ocfs2_remove_from_cache (INODE_CACHE (main_bm_inode), group_bh);brelse (group_bh); out_unlock:
Signature · 1 line
2.43.0
Joseph Qi
maintainerQuoted 30 lines · ZhengYuan Huang
[BUG]
OCFS2_IOC_GROUP_ADDcan trigger aBUG_ONinocfs2_set_new_buffer_uptodate():kernel BUG at
fs/ocfs2/uptodate.c:509!Oops: invalid opcode: 0000 [#1] SMP KASAN NOPTI RIP: 0010:ocfs2_set_new_buffer_uptodate+0x194/0x1e0 fs/ocfs2/uptodate.c:509 Code: ffffe88f 42b9fe4c 89e64889 dfe8b4df Call Trace: ocfs2_group_add+0x3f1/0x1510 fs/ocfs2/resize.c:507 ocfs2_ioctl+0x309/0x6e0 fs/ocfs2/ioctl.c:887 vfs_ioctl fs/ioctl.c:51 [inline] __do_sys_ioctl fs/ioctl.c:597 [inline] __se_sys_ioctl fs/ioctl.c:583 [inline] __x64_sys_ioctl+0x197/0x1e0 fs/ioctl.c:583 x64_sys_call+0x1144/0x26a0 arch/x86/include/generated/asm/syscalls_64.h:17 do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline] do_syscall_64+0x93/0xf80 arch/x86/entry/syscall_64.c:94 entry_SYSCALL_64_after_hwframe+0x76/0x7e RIP: 0033:0x7bbfb55a966d[CAUSE]
ocfs2_group_add()callsocfs2_set_new_buffer_uptodate()on a user-controlled group block beforeocfs2_verify_group_and_input()validates that block number. That helper is only valid for newly allocated metadata and asserts that the block is not already present in the chosen metadata cache. The code also usesINODE_CACHE(inode)even though the group descriptor belongs tomain_bm_inodeand later journal accesses use that cache context instead.
Quoted 8 lines
[FIX]
Validate the on-disk group descriptor before caching it, then add it to the metadata cache tracked byINODE_CACHE(main_bm_inode). Update the error path to remove the buffer from the same cache context so the group buffer lifetime stays consistent across validation, journaling, and cleanup.Signed-off-by ZhengYuan Huang <[email protected]>
Missing Fixes tag:
Quoted 9 lines
--- fs/ocfs2/resize.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)fs/ocfs2/resize.c +0
-0
fs/ocfs2/resize.c int ocfs2_group_add(struct inode *inode, struct ocfs2_new_group_input *input) · line 511
Quoted 10 lines
fs/ocfs2/resize.c +0
-2@@ -504,14 +504,14 @@ int ocfs2_group_add(struct inode *inode, struct ocfs2_new_group_input *input) goto out_unlock; } -ocfs2_set_new_buffer_uptodate (INODE_CACHE (inode), group_bh); - ret =ocfs2_verify_group_and_input (main_bm_inode, fe, input, group_bh);if (ret) {mlog_errno (ret); goto out_free_group_bh; }
Since ocfs2_set_new_buffer_uptodate is now moved down, the error goto above should also be changed.
e.g. do the same ioctl twice, the second will remove the cache which should be valid. This is not the expected behavior.
Thanks,
Joseph
Quoted 15 lines
fs/ocfs2/resize.c +3
-1+ocfs2_set_new_buffer_uptodate (INODE_CACHE (main_bm_inode), group_bh); +trace_ocfs2_group_add ((unsigned long long )input->group, input->chain, input->clusters, input->frees);@@ -575,7 +575,7 @@ int ocfs2_group_add(struct inode *inode, struct ocfs2_new_group_input *input) out_free_group_bh:if (ret <0 ) -ocfs2_remove_from_cache (INODE_CACHE (inode), group_bh); +ocfs2_remove_from_cache (INODE_CACHE (main_bm_inode), group_bh);brelse (group_bh); out_unlock:
Quoted 26 lines · Joseph Qi
Missing Fixes tag:
Fixes 7909f2bf8353 ("[PATCH 2/2] ocfs2: Implement group add for online resize")Quoted 9 lines
--- fs/ocfs2/resize.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)fs/ocfs2/resize.c +0
-0fs/ocfs2/resize.c int ocfs2_group_add(struct inode *inode, struct ocfs2_new_group_input *input) · line 511
Quoted 10 lines
fs/ocfs2/resize.c +0
-2@@ -504,14 +504,14 @@ int ocfs2_group_add(struct inode *inode, struct ocfs2_new_group_input *input) goto out_unlock; } -ocfs2_set_new_buffer_uptodate (INODE_CACHE (inode), group_bh); - ret =ocfs2_verify_group_and_input (main_bm_inode, fe, input, group_bh);if (ret) {mlog_errno (ret); goto out_free_group_bh; }Since
ocfs2_set_new_buffer_uptodateis now moved down, the error goto above should also be changed.
Quoted 5 lines
e.g. do the same ioctl twice, the second will remove the cache which should be valid. This is not the expected behavior.
Thanks,
Joseph
Thanks for pointing this out.
The issue is valid, and I’ve addressed it in v2 by fixing the error path accordingly.
v2 patch has been sent.
Thanks,
ZhengYuan Huang