erofs: Fix pcluster memleak when its block address is zero
PATCH v3 1/1 erofs: Fix pcluster memleak when its block address is zero
syzkaller reported a memleak:
https://syzkaller.appspot.com/bug?id=62f37ff612f0021641eda5b17f056f1668aa9aed
unreferenced object 0xffff88811009c7f8 (size 136):
...
backtrace:
[<ffffffff821db19b>] z_erofs_do_read_page+0x99b/0x1740
[<ffffffff821dee9e>] z_erofs_readahead+0x24e/0x580
[<ffffffff814bc0d6>] read_pages+0x86/0x3d0
...syzkaller constructed a case: in z_erofs_register_pcluster(), ztailpacking = false and map->m_pa = zero. This makes pcl->obj.index be zero although pcl is not a inline pcluster.
Then following path adds refcount for grp, but the refcount won't be put because pcl is inline.
z_erofs_readahead()
z_erofs_do_read_page() # for another page
z_erofs_collector_begin()
erofs_find_workgroup()
erofs_workgroup_get()
Since it's illegal for the block address of a pcluster to be zero, add check here to avoid registering the pcluster which would be leaked.
---
v1 -> v2:
As Gao's advice, we should fail to register pcluster if m_pa is zero. Maked it this way and changed the commit message.
v2 -> v3:
Slightly fix commit message and add -next tag.
--- fs/erofs/zdata.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
fs/erofs/zdata.c +2 -1
@@ -488,7 +488,8 @@ static int z_erofs_register_pcluster(struct z_erofs_decompress_frontend *fe)struct erofs_workgroup *grp;int err; -if (!(map->m_flags &EROFS_MAP_ENCODED )) { +if (!(map->m_flags &EROFS_MAP_ENCODED ) || + !(map->m_pa >>PAGE_SHIFT )) {DBG_BUGON (1 );return -EFSCORRUPTED ; }
Signature · 1 line
2.17.1
Quoted 24 lines · Chen Zhongjin
syzkaller reported a memleak:
https://syzkaller.appspot.com/bug?id=62f37ff612f0021641eda5b17f056f1668aa9aedunreferenced object
0xffff88811009c7f8(size 136):
...
backtrace:[<ffffffff821db19b>] z_erofs_do_read_page+0x99b/0x1740 [<ffffffff821dee9e>] z_erofs_readahead+0x24e/0x580 [<ffffffff814bc0d6>] read_pages+0x86/0x3d0 ...syzkaller constructed a case: in
z_erofs_register_pcluster(), ztailpacking = false andmap->m_pa= zero. This makespcl->obj.indexbe zero although pcl is not a inline pcluster.Then following path adds refcount for grp, but the refcount won't be put because pcl is inline.
z_erofs_readahead()
z_erofs_do_read_page()# for another page
z_erofs_collector_begin()
erofs_find_workgroup()
erofs_workgroup_get()
Quoted 6 lines
Since it's illegal for the block address of a pcluster to be zero, add check here to avoid registering the pcluster which would be leaked.
Fixes cecf864d3d76 ("erofs: support inline data decompression")Reported-by [email protected]Signed-off-by Chen Zhongjin <[email protected]>
Thanks,
Gao Xiang
Quoted 24 lines · On Mon, 5 Dec 2022 11:49:57 +0800 Chen Zhongjin via Linux-erofs…
syzkaller reported a memleak:
https://syzkaller.appspot.com/bug?id=62f37ff612f0021641eda5b17f056f1668aa9aedunreferenced object
0xffff88811009c7f8(size 136):
...
backtrace:[<ffffffff821db19b>] z_erofs_do_read_page+0x99b/0x1740 [<ffffffff821dee9e>] z_erofs_readahead+0x24e/0x580 [<ffffffff814bc0d6>] read_pages+0x86/0x3d0 ...syzkaller constructed a case: in
z_erofs_register_pcluster(), ztailpacking = false andmap->m_pa= zero. This makespcl->obj.indexbe zero although pcl is not a inline pcluster.Then following path adds refcount for grp, but the refcount won't be put because pcl is inline.
z_erofs_readahead()
z_erofs_do_read_page()# for another page
z_erofs_collector_begin()
erofs_find_workgroup()
erofs_workgroup_get()
Quoted 6 lines
Since it's illegal for the block address of a pcluster to be zero, add check here to avoid registering the pcluster which would be leaked.
Fixes cecf864d3d76 ("erofs: support inline data decompression")Reported-by [email protected]Signed-off-by Chen Zhongjin <[email protected]>
Thanks.
Quoted 25 lines
---
v1 -> v2:
As Gao's advice, we should fail to register pcluster ifm_pais zero. Maked it this way and changed the commit message.v2 -> v3:
Slightly fix commit message and add -next tag.--- fs/erofs/zdata.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)fs/erofs/zdata.c +2
-1@@ -488,7 +488,8 @@ static int z_erofs_register_pcluster(struct z_erofs_decompress_frontend *fe)struct erofs_workgroup *grp;int err; -if (!(map->m_flags &EROFS_MAP_ENCODED )) { +if (!(map->m_flags &EROFS_MAP_ENCODED ) || + !(map->m_pa >>PAGE_SHIFT )) {DBG_BUGON (1 );return -EFSCORRUPTED ; }
Hi all,
Quoted 30 lines · Chen Zhongjin
syzkaller reported a memleak:
https://syzkaller.appspot.com/bug?id=62f37ff612f0021641eda5b17f056f1668aa9aedunreferenced object
0xffff88811009c7f8(size 136):
...
backtrace:[<ffffffff821db19b>] z_erofs_do_read_page+0x99b/0x1740 [<ffffffff821dee9e>] z_erofs_readahead+0x24e/0x580 [<ffffffff814bc0d6>] read_pages+0x86/0x3d0 ...syzkaller constructed a case: in
z_erofs_register_pcluster(), ztailpacking = false andmap->m_pa= zero. This makespcl->obj.indexbe zero although pcl is not a inline pcluster.Then following path adds refcount for grp, but the refcount won't be put because pcl is inline.
z_erofs_readahead()
z_erofs_do_read_page()# for another page
z_erofs_collector_begin()
erofs_find_workgroup()
erofs_workgroup_get()Since it's illegal for the block address of a pcluster to be zero, add check here to avoid registering the pcluster which would be leaked.
Fixes cecf864d3d76 ("erofs: support inline data decompression")Reported-by [email protected]Signed-off-by Chen Zhongjin <[email protected]>
Quoted 7 lines
---
v1 -> v2:
As Gao's advice, we should fail to register pcluster ifm_pais zero. Maked it this way and changed the commit message.v2 -> v3:
Slightly fix commit message and add -next tag.
I've updated the patch itself as below
(Since we only need to fail out for non-tailpacking cases, tailpacking
inline inodes could still have m_pa < EROFS_BLKSIZ):
From f5e037e760d338ca0c116e507be663cb843d42f0 Mon Sep 17 00:00:00 2001
From: Chen Zhongjin <[email protected]>
Date: Mon, 5 Dec 2022 11:49:57 +0800
Subject: [PATCH] erofs: Fix pcluster memleak when its block address is zero
syzkaller reported a memleak:
https://syzkaller.appspot.com/bug?id=62f37ff612f0021641eda5b17f056f1668aa9aed
unreferenced object 0xffff88811009c7f8 (size 136):
...
backtrace:
[<ffffffff821db19b>] z_erofs_do_read_page+0x99b/0x1740
[<ffffffff821dee9e>] z_erofs_readahead+0x24e/0x580
[<ffffffff814bc0d6>] read_pages+0x86/0x3d0
...syzkaller constructed a case: in z_erofs_register_pcluster(), ztailpacking = false and map->m_pa = zero. This makes pcl->obj.index be zero although pcl is not a inline pcluster.
Then following path adds refcount for grp, but the refcount won't be put because pcl is inline.
z_erofs_readahead()
z_erofs_do_read_page() # for another page
z_erofs_collector_begin()
erofs_find_workgroup()
erofs_workgroup_get()
Since it's illegal for the block address of a non-inlined pcluster to be zero, add check here to avoid registering the pcluster which would be leaked.
--- fs/erofs/zdata.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
fs/erofs/zdata.c +2 -1
@@ -496,7 +496,8 @@ static int z_erofs_register_pcluster(struct z_erofs_decompress_frontend *fe)struct erofs_workgroup *grp;int err; -if (!(map->m_flags &EROFS_MAP_ENCODED )) { +if (!(map->m_flags &EROFS_MAP_ENCODED ) || + (!ztailpacking && !(map->m_pa >>PAGE_SHIFT ))) {DBG_BUGON (1 );return -EFSCORRUPTED ; }
Signature · 1 line
2.30.2
Quoted 78 lines · Gao Xiang
Hi all,
Quoted 30 lines · Chen Zhongjin
syzkaller reported a memleak:
https://syzkaller.appspot.com/bug?id=62f37ff612f0021641eda5b17f056f1668aa9aedunreferenced object
0xffff88811009c7f8(size 136):... backtrace: [<ffffffff821db19b>] z_erofs_do_read_page+0x99b/0x1740 [<ffffffff821dee9e>] z_erofs_readahead+0x24e/0x580 [<ffffffff814bc0d6>] read_pages+0x86/0x3d0 ...syzkaller constructed a case: in
z_erofs_register_pcluster(), ztailpacking = false andmap->m_pa= zero. This makespcl->obj.indexbe zero although pcl is not a inline pcluster.Then following path adds refcount for grp, but the refcount won't be put because pcl is inline.
z_erofs_readahead()
z_erofs_do_read_page()# for another page
z_erofs_collector_begin()
erofs_find_workgroup()
erofs_workgroup_get()Since it's illegal for the block address of a pcluster to be zero, add check here to avoid registering the pcluster which would be leaked.
Fixes cecf864d3d76 ("erofs: support inline data decompression")Reported-by [email protected]Signed-off-by Chen Zhongjin <[email protected]>Quoted 7 lines
---
v1 -> v2:
As Gao's advice, we should fail to register pcluster ifm_pais zero. Maked it this way and changed the commit message.v2 -> v3:
Slightly fix commit message and add -next tag.I've updated the patch itself as below
(Since we only need to fail out for non-tailpacking cases, tailpacking
inline inodes could still havem_pa<EROFS_BLKSIZ):From
f5e037e760d338ca0c116e507be663cb843d42f0Mon Sep 17 00:00:00 2001
From: Chen Zhongjin <[email protected]>
Date: Mon, 5 Dec 2022 11:49:57 +0800
Subject: [PATCH] erofs: Fix pcluster memleak when its block address is zerosyzkaller reported a memleak:
https://syzkaller.appspot.com/bug?id=62f37ff612f0021641eda5b17f056f1668aa9aedunreferenced object
0xffff88811009c7f8(size 136):... backtrace: [<ffffffff821db19b>] z_erofs_do_read_page+0x99b/0x1740 [<ffffffff821dee9e>] z_erofs_readahead+0x24e/0x580 [<ffffffff814bc0d6>] read_pages+0x86/0x3d0 ...syzkaller constructed a case: in
z_erofs_register_pcluster(), ztailpacking = false andmap->m_pa= zero. This makespcl->obj.indexbe zero although pcl is not a inline pcluster.Then following path adds refcount for grp, but the refcount won't be put because pcl is inline.
z_erofs_readahead()
z_erofs_do_read_page()# for another page
z_erofs_collector_begin()
erofs_find_workgroup()
erofs_workgroup_get()Since it's illegal for the block address of a non-inlined pcluster to be zero, add check here to avoid registering the pcluster which would be leaked.
Quoted 6 lines
Fixes cecf864d3d76 ("erofs: support inline data decompression")Reported-by [email protected]Signed-off-by Chen Zhongjin <[email protected]>Reviewed-by Yue Hu <[email protected]>Reviewed-by Gao Xiang <[email protected]>Signed-off-by Gao Xiang <[email protected]>
Thanks,