Message ID | 1530598800-9174-1-git-send-email-gujx@cn.fujitsu.com (mailing list archive) |
---|---|
State | New, archived |
Headers | show |
On 2018年07月03日 14:20, Gu Jinxiang wrote: > Reported in https://bugzilla.kernel.org/show_bug.cgi?id=199839, > which has a invalid chunk, not return error opportunlly. > > Add chunk type check in btrfs_check_chunk_valid, > to make error be returned in advance. > > Reported-by: Xu Wen <wen.xu@gatech.edu> > Signed-off-by: Gu Jinxiang <gujx@cn.fujitsu.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Thanks, Qu > --- > changelog: > v2: > deal with comment by Qu, add precise check for chunk type. > > fs/btrfs/volumes.c | 27 +++++++++++++++++++++++++++ > 1 file changed, 27 insertions(+) > > diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c > index e034ad9e23b4..fcdb1a75d653 100644 > --- a/fs/btrfs/volumes.c > +++ b/fs/btrfs/volumes.c > @@ -6401,6 +6401,8 @@ static int btrfs_check_chunk_valid(struct btrfs_fs_info *fs_info, > u16 num_stripes; > u16 sub_stripes; > u64 type; > + u64 features; > + int mixed = 0; > > length = btrfs_chunk_length(leaf, chunk); > stripe_len = btrfs_chunk_stripe_len(leaf, chunk); > @@ -6439,6 +6441,31 @@ static int btrfs_check_chunk_valid(struct btrfs_fs_info *fs_info, > btrfs_chunk_type(leaf, chunk)); > return -EIO; > } > + > + if (!(type & BTRFS_BLOCK_GROUP_TYPE_MASK)) { > + btrfs_err(fs_info, "missing chunk type flag: %llu", type); > + return -EIO; > + } > + > + if ((type & BTRFS_BLOCK_GROUP_SYSTEM) && > + (type & (BTRFS_BLOCK_GROUP_METADATA | BTRFS_BLOCK_GROUP_DATA))) { > + btrfs_err(fs_info, "duplicate chunk type flag: %llu", type); > + return -EIO; > + } > + > + features = btrfs_super_incompat_flags(fs_info->super_copy); > + if (features & BTRFS_FEATURE_INCOMPAT_MIXED_GROUPS) > + mixed = 1; > + > + if (!mixed) { > + if (type & > + (BTRFS_BLOCK_GROUP_METADATA & BTRFS_BLOCK_GROUP_DATA)) { > + btrfs_err(fs_info, > + "mixed chunk type flag in non-mixed mode: %llu", type); > + return -EIO; > + } > + } > + > if ((type & BTRFS_BLOCK_GROUP_RAID10 && sub_stripes != 2) || > (type & BTRFS_BLOCK_GROUP_RAID1 && num_stripes < 1) || > (type & BTRFS_BLOCK_GROUP_RAID5 && num_stripes < 2) || >
On 3.07.2018 09:20, Gu Jinxiang wrote: > Reported in https://bugzilla.kernel.org/show_bug.cgi?id=199839, > which has a invalid chunk, not return error opportunlly. > > Add chunk type check in btrfs_check_chunk_valid, > to make error be returned in advance. > > Reported-by: Xu Wen <wen.xu@gatech.edu> > Signed-off-by: Gu Jinxiang <gujx@cn.fujitsu.com> > --- > changelog: > v2: > deal with comment by Qu, add precise check for chunk type. > > fs/btrfs/volumes.c | 27 +++++++++++++++++++++++++++ > 1 file changed, 27 insertions(+) > > diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c > index e034ad9e23b4..fcdb1a75d653 100644 > --- a/fs/btrfs/volumes.c > +++ b/fs/btrfs/volumes.c > @@ -6401,6 +6401,8 @@ static int btrfs_check_chunk_valid(struct btrfs_fs_info *fs_info, > u16 num_stripes; > u16 sub_stripes; > u64 type; > + u64 features; > + int mixed = 0; > > length = btrfs_chunk_length(leaf, chunk); > stripe_len = btrfs_chunk_stripe_len(leaf, chunk); > @@ -6439,6 +6441,31 @@ static int btrfs_check_chunk_valid(struct btrfs_fs_info *fs_info, > btrfs_chunk_type(leaf, chunk)); > return -EIO; > } > + > + if (!(type & BTRFS_BLOCK_GROUP_TYPE_MASK)) { > + btrfs_err(fs_info, "missing chunk type flag: %llu", type); > + return -EIO; > + } > + > + if ((type & BTRFS_BLOCK_GROUP_SYSTEM) && > + (type & (BTRFS_BLOCK_GROUP_METADATA | BTRFS_BLOCK_GROUP_DATA))) { > + btrfs_err(fs_info, "duplicate chunk type flag: %llu", type); This is not really a duplicate chunk type flag but instead an invalid combination. SYSTEM is always a dedicated chunk, rephrase the error message. > + return -EIO; > + } > + > + features = btrfs_super_incompat_flags(fs_info->super_copy); > + if (features & BTRFS_FEATURE_INCOMPAT_MIXED_GROUPS) > + mixed = 1; > + > + if (!mixed) { > + if (type & > + (BTRFS_BLOCK_GROUP_METADATA & BTRFS_BLOCK_GROUP_DATA)) { > + btrfs_err(fs_info, > + "mixed chunk type flag in non-mixed mode: %llu", type); > + return -EIO; > + } > + } > + > if ((type & BTRFS_BLOCK_GROUP_RAID10 && sub_stripes != 2) || > (type & BTRFS_BLOCK_GROUP_RAID1 && num_stripes < 1) || > (type & BTRFS_BLOCK_GROUP_RAID5 && num_stripes < 2) || > -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
> -----Original Message----- > From: Nikolay Borisov [mailto:nborisov@suse.com] > Sent: Wednesday, July 04, 2018 4:15 PM > To: Gu, Jinxiang/顾 金香 <gujx@cn.fujitsu.com>; linux-btrfs@vger.kernel.org > Cc: quwenruo.btrfs@gmx.com; wen.xu@gatech.edu; xuwen.sjtu@gmail.com > Subject: Re: [PATCH v2] btrfs: Add chunk type check in read a chunk > > > > On 3.07.2018 09:20, Gu Jinxiang wrote: > > Reported in https://bugzilla.kernel.org/show_bug.cgi?id=199839, > > which has a invalid chunk, not return error opportunlly. > > > > Add chunk type check in btrfs_check_chunk_valid, > > to make error be returned in advance. > > > > Reported-by: Xu Wen <wen.xu@gatech.edu> > > Signed-off-by: Gu Jinxiang <gujx@cn.fujitsu.com> > > --- > > changelog: > > v2: > > deal with comment by Qu, add precise check for chunk type. > > > > fs/btrfs/volumes.c | 27 +++++++++++++++++++++++++++ > > 1 file changed, 27 insertions(+) > > > > diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c > > index e034ad9e23b4..fcdb1a75d653 100644 > > --- a/fs/btrfs/volumes.c > > +++ b/fs/btrfs/volumes.c > > @@ -6401,6 +6401,8 @@ static int btrfs_check_chunk_valid(struct btrfs_fs_info *fs_info, > > u16 num_stripes; > > u16 sub_stripes; > > u64 type; > > + u64 features; > > + int mixed = 0; > > > > length = btrfs_chunk_length(leaf, chunk); > > stripe_len = btrfs_chunk_stripe_len(leaf, chunk); > > @@ -6439,6 +6441,31 @@ static int btrfs_check_chunk_valid(struct btrfs_fs_info *fs_info, > > btrfs_chunk_type(leaf, chunk)); > > return -EIO; > > } > > + > > + if (!(type & BTRFS_BLOCK_GROUP_TYPE_MASK)) { > > + btrfs_err(fs_info, "missing chunk type flag: %llu", type); > > + return -EIO; > > + } > > + > > + if ((type & BTRFS_BLOCK_GROUP_SYSTEM) && > > + (type & (BTRFS_BLOCK_GROUP_METADATA | BTRFS_BLOCK_GROUP_DATA))) { > > + btrfs_err(fs_info, "duplicate chunk type flag: %llu", type); > > This is not really a duplicate chunk type flag but instead an invalid > combination. SYSTEM is always a dedicated chunk, rephrase the error > message. Yes. This message should be updated. But [PATCH 1/5] btrfs: tree-checker: Verify block_group_item sent by Qu, checks the chunk type in tree checker, which be called when read a bio. So this patch should not be merged. > > > + return -EIO; > > + } > > + > > + features = btrfs_super_incompat_flags(fs_info->super_copy); > > + if (features & BTRFS_FEATURE_INCOMPAT_MIXED_GROUPS) > > + mixed = 1; > > + > > + if (!mixed) { > > + if (type & > > + (BTRFS_BLOCK_GROUP_METADATA & BTRFS_BLOCK_GROUP_DATA)) { > > + btrfs_err(fs_info, > > + "mixed chunk type flag in non-mixed mode: %llu", type); > > + return -EIO; > > + } > > + } > > + > > if ((type & BTRFS_BLOCK_GROUP_RAID10 && sub_stripes != 2) || > > (type & BTRFS_BLOCK_GROUP_RAID1 && num_stripes < 1) || > > (type & BTRFS_BLOCK_GROUP_RAID5 && num_stripes < 2) || > > >
On 2018年07月04日 16:39, Gu, Jinxiang wrote: > > >> -----Original Message----- >> From: Nikolay Borisov [mailto:nborisov@suse.com] >> Sent: Wednesday, July 04, 2018 4:15 PM >> To: Gu, Jinxiang/顾 金香 <gujx@cn.fujitsu.com>; linux-btrfs@vger.kernel.org >> Cc: quwenruo.btrfs@gmx.com; wen.xu@gatech.edu; xuwen.sjtu@gmail.com >> Subject: Re: [PATCH v2] btrfs: Add chunk type check in read a chunk >> >> >> >> On 3.07.2018 09:20, Gu Jinxiang wrote: >>> Reported in https://bugzilla.kernel.org/show_bug.cgi?id=199839, >>> which has a invalid chunk, not return error opportunlly. >>> >>> Add chunk type check in btrfs_check_chunk_valid, >>> to make error be returned in advance. >>> >>> Reported-by: Xu Wen <wen.xu@gatech.edu> >>> Signed-off-by: Gu Jinxiang <gujx@cn.fujitsu.com> >>> --- >>> changelog: >>> v2: >>> deal with comment by Qu, add precise check for chunk type. >>> >>> fs/btrfs/volumes.c | 27 +++++++++++++++++++++++++++ >>> 1 file changed, 27 insertions(+) >>> >>> diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c >>> index e034ad9e23b4..fcdb1a75d653 100644 >>> --- a/fs/btrfs/volumes.c >>> +++ b/fs/btrfs/volumes.c >>> @@ -6401,6 +6401,8 @@ static int btrfs_check_chunk_valid(struct btrfs_fs_info *fs_info, >>> u16 num_stripes; >>> u16 sub_stripes; >>> u64 type; >>> + u64 features; >>> + int mixed = 0; >>> >>> length = btrfs_chunk_length(leaf, chunk); >>> stripe_len = btrfs_chunk_stripe_len(leaf, chunk); >>> @@ -6439,6 +6441,31 @@ static int btrfs_check_chunk_valid(struct btrfs_fs_info *fs_info, >>> btrfs_chunk_type(leaf, chunk)); >>> return -EIO; >>> } >>> + >>> + if (!(type & BTRFS_BLOCK_GROUP_TYPE_MASK)) { >>> + btrfs_err(fs_info, "missing chunk type flag: %llu", type); >>> + return -EIO; >>> + } >>> + >>> + if ((type & BTRFS_BLOCK_GROUP_SYSTEM) && >>> + (type & (BTRFS_BLOCK_GROUP_METADATA | BTRFS_BLOCK_GROUP_DATA))) { >>> + btrfs_err(fs_info, "duplicate chunk type flag: %llu", type); >> >> This is not really a duplicate chunk type flag but instead an invalid >> combination. SYSTEM is always a dedicated chunk, rephrase the error >> message. > > Yes. This message should be updated. > But [PATCH 1/5] btrfs: tree-checker: Verify block_group_item sent by Qu, checks the chunk type > in tree checker, which be called when read a bio. So this patch should not be merged. Nope, this patch is still needed. Although I added block group -> chunk checker, and it indeed detects all the crafted images, doing chunk verification is nothing wrong. If chunk verification can detect anything wrong, we can error out even earlier. So I still think this patch is needed. (Although I'd like to move a lot of chunk verification to tree-checker later) Thanks, Qu > >> >>> + return -EIO; >>> + } >>> + >>> + features = btrfs_super_incompat_flags(fs_info->super_copy); >>> + if (features & BTRFS_FEATURE_INCOMPAT_MIXED_GROUPS) >>> + mixed = 1; >>> + >>> + if (!mixed) { >>> + if (type & >>> + (BTRFS_BLOCK_GROUP_METADATA & BTRFS_BLOCK_GROUP_DATA)) { >>> + btrfs_err(fs_info, >>> + "mixed chunk type flag in non-mixed mode: %llu", type); >>> + return -EIO; >>> + } >>> + } >>> + >>> if ((type & BTRFS_BLOCK_GROUP_RAID10 && sub_stripes != 2) || >>> (type & BTRFS_BLOCK_GROUP_RAID1 && num_stripes < 1) || >>> (type & BTRFS_BLOCK_GROUP_RAID5 && num_stripes < 2) || >>> >> > > > > N�����r��y���b�X��ǧv�^�){.n�+����{�n�߲)���w*jg��������ݢj/���z�ޖ��2�ޙ���&�)ߡ�a�����G���h��j:+v���w�٥ > -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Wed, Jul 04, 2018 at 05:49:50PM +0800, Qu Wenruo wrote: > On 2018年07月04日 16:39, Gu, Jinxiang wrote: > >> -----Original Message----- > >> From: Nikolay Borisov [mailto:nborisov@suse.com] > >> Sent: Wednesday, July 04, 2018 4:15 PM > >> To: Gu, Jinxiang/顾 金香 <gujx@cn.fujitsu.com>; linux-btrfs@vger.kernel.org > >> Cc: quwenruo.btrfs@gmx.com; wen.xu@gatech.edu; xuwen.sjtu@gmail.com > >> Subject: Re: [PATCH v2] btrfs: Add chunk type check in read a chunk > >> On 3.07.2018 09:20, Gu Jinxiang wrote: > >>> Reported in https://bugzilla.kernel.org/show_bug.cgi?id=199839, > >>> which has a invalid chunk, not return error opportunlly. > >>> > >>> Add chunk type check in btrfs_check_chunk_valid, > >>> to make error be returned in advance. > >>> > >>> Reported-by: Xu Wen <wen.xu@gatech.edu> > >>> Signed-off-by: Gu Jinxiang <gujx@cn.fujitsu.com> > >>> --- > >>> changelog: > >>> v2: > >>> deal with comment by Qu, add precise check for chunk type. > >>> > >>> fs/btrfs/volumes.c | 27 +++++++++++++++++++++++++++ > >>> 1 file changed, 27 insertions(+) > >>> > >>> diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c > >>> index e034ad9e23b4..fcdb1a75d653 100644 > >>> --- a/fs/btrfs/volumes.c > >>> +++ b/fs/btrfs/volumes.c > >>> @@ -6401,6 +6401,8 @@ static int btrfs_check_chunk_valid(struct btrfs_fs_info *fs_info, > >>> u16 num_stripes; > >>> u16 sub_stripes; > >>> u64 type; > >>> + u64 features; > >>> + int mixed = 0; > >>> > >>> length = btrfs_chunk_length(leaf, chunk); > >>> stripe_len = btrfs_chunk_stripe_len(leaf, chunk); > >>> @@ -6439,6 +6441,31 @@ static int btrfs_check_chunk_valid(struct btrfs_fs_info *fs_info, > >>> btrfs_chunk_type(leaf, chunk)); > >>> return -EIO; > >>> } > >>> + > >>> + if (!(type & BTRFS_BLOCK_GROUP_TYPE_MASK)) { > >>> + btrfs_err(fs_info, "missing chunk type flag: %llu", type); > >>> + return -EIO; > >>> + } > >>> + > >>> + if ((type & BTRFS_BLOCK_GROUP_SYSTEM) && > >>> + (type & (BTRFS_BLOCK_GROUP_METADATA | BTRFS_BLOCK_GROUP_DATA))) { > >>> + btrfs_err(fs_info, "duplicate chunk type flag: %llu", type); > >> > >> This is not really a duplicate chunk type flag but instead an invalid > >> combination. SYSTEM is always a dedicated chunk, rephrase the error > >> message. > > > > Yes. This message should be updated. > > But [PATCH 1/5] btrfs: tree-checker: Verify block_group_item sent by Qu, checks the chunk type > > in tree checker, which be called when read a bio. So this patch should not be merged. > > Nope, this patch is still needed. > > Although I added block group -> chunk checker, and it indeed detects all > the crafted images, doing chunk verification is nothing wrong. > > If chunk verification can detect anything wrong, we can error out even > earlier. > > So I still think this patch is needed. > (Although I'd like to move a lot of chunk verification to tree-checker > later) I'd rather add that to the tree-checker.c now and the enhanced checks. btrfs_check_chunk_valid is self-contained and does not use anything that's local to volumes.c. -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index e034ad9e23b4..fcdb1a75d653 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -6401,6 +6401,8 @@ static int btrfs_check_chunk_valid(struct btrfs_fs_info *fs_info, u16 num_stripes; u16 sub_stripes; u64 type; + u64 features; + int mixed = 0; length = btrfs_chunk_length(leaf, chunk); stripe_len = btrfs_chunk_stripe_len(leaf, chunk); @@ -6439,6 +6441,31 @@ static int btrfs_check_chunk_valid(struct btrfs_fs_info *fs_info, btrfs_chunk_type(leaf, chunk)); return -EIO; } + + if (!(type & BTRFS_BLOCK_GROUP_TYPE_MASK)) { + btrfs_err(fs_info, "missing chunk type flag: %llu", type); + return -EIO; + } + + if ((type & BTRFS_BLOCK_GROUP_SYSTEM) && + (type & (BTRFS_BLOCK_GROUP_METADATA | BTRFS_BLOCK_GROUP_DATA))) { + btrfs_err(fs_info, "duplicate chunk type flag: %llu", type); + return -EIO; + } + + features = btrfs_super_incompat_flags(fs_info->super_copy); + if (features & BTRFS_FEATURE_INCOMPAT_MIXED_GROUPS) + mixed = 1; + + if (!mixed) { + if (type & + (BTRFS_BLOCK_GROUP_METADATA & BTRFS_BLOCK_GROUP_DATA)) { + btrfs_err(fs_info, + "mixed chunk type flag in non-mixed mode: %llu", type); + return -EIO; + } + } + if ((type & BTRFS_BLOCK_GROUP_RAID10 && sub_stripes != 2) || (type & BTRFS_BLOCK_GROUP_RAID1 && num_stripes < 1) || (type & BTRFS_BLOCK_GROUP_RAID5 && num_stripes < 2) ||
Reported in https://bugzilla.kernel.org/show_bug.cgi?id=199839, which has a invalid chunk, not return error opportunlly. Add chunk type check in btrfs_check_chunk_valid, to make error be returned in advance. Reported-by: Xu Wen <wen.xu@gatech.edu> Signed-off-by: Gu Jinxiang <gujx@cn.fujitsu.com> --- changelog: v2: deal with comment by Qu, add precise check for chunk type. fs/btrfs/volumes.c | 27 +++++++++++++++++++++++++++ 1 file changed, 27 insertions(+)