[PATCH 6.5.y] erofs: ensure that the post-EOF tails are all zeroed

[PATCH 6.5.y] erofs: ensure that the post-EOF tails are all zeroed

[Gao Xiang]

commit e4c1cf523d820730a86cae2c6d55924833b6f7ac upstream.

This was accidentally fixed up in commit e4c1cf523d82 but we can't
take the full change due to other dependancy issues, so here is just
the actual bugfix that is needed.

[Background]

keltargw reported an issue [1] that with mmaped I/Os, sometimes the
tail of the last page (after file ends) is not filled with zeroes.

The root cause is that such tail page could be wrongly selected for
inplace I/Os so the zeroed part will then be filled with compressed
data instead of zeroes.

A simple fix is to avoid doing inplace I/Os for such tail parts,
actually that was already fixed upstream in commit e4c1cf523d82
("erofs: tidy up z_erofs_do_read_page()") by accident.

[1] https://lore.kernel.org/r/3ad8b469-25db-a297-21f9-75db2d6ad224@linux.alibaba.com

Reported-by: keltargw <keltar.gw@gmail.com>
Fixes: 3883a79abd02 ("staging: erofs: introduce VLE decompression support")
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
---
 fs/erofs/zdata.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/fs/erofs/zdata.c b/fs/erofs/zdata.c
index de4f12152b62..9c9350eb1704 100644
--- a/fs/erofs/zdata.c
+++ b/fs/erofs/zdata.c
@@ -1038,6 +1038,8 @@ static int z_erofs_do_read_page(struct z_erofs_decompress_frontend *fe,
 	cur = end - min_t(erofs_off_t, offset + end - map->m_la, end);
 	if (!(map->m_flags & EROFS_MAP_MAPPED)) {
 		zero_user_segment(page, cur, end);
+		++spiltted;
+		tight = false;
 		goto next_part;
 	}
 	if (map->m_flags & EROFS_MAP_FRAGMENT) {
-- 
2.24.4

[PATCH 6.4.y] erofs: ensure that the post-EOF tails are all zeroed

[Gao Xiang]

commit e4c1cf523d820730a86cae2c6d55924833b6f7ac upstream.

This was accidentally fixed up in commit e4c1cf523d82 but we can't
take the full change due to other dependancy issues, so here is just
the actual bugfix that is needed.

[Background]

keltargw reported an issue [1] that with mmaped I/Os, sometimes the
tail of the last page (after file ends) is not filled with zeroes.

The root cause is that such tail page could be wrongly selected for
inplace I/Os so the zeroed part will then be filled with compressed
data instead of zeroes.

A simple fix is to avoid doing inplace I/Os for such tail parts,
actually that was already fixed upstream in commit e4c1cf523d82
("erofs: tidy up z_erofs_do_read_page()") by accident.

[1] https://lore.kernel.org/r/3ad8b469-25db-a297-21f9-75db2d6ad224@linux.alibaba.com

Reported-by: keltargw <keltar.gw@gmail.com>
Fixes: 3883a79abd02 ("staging: erofs: introduce VLE decompression support")
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
---
 fs/erofs/zdata.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/fs/erofs/zdata.c b/fs/erofs/zdata.c
index 470988bb7867..9a7c8bb0590f 100644
--- a/fs/erofs/zdata.c
+++ b/fs/erofs/zdata.c
@@ -993,6 +993,8 @@ static int z_erofs_do_read_page(struct z_erofs_decompress_frontend *fe,
 	cur = end - min_t(erofs_off_t, offset + end - map->m_la, end);
 	if (!(map->m_flags & EROFS_MAP_MAPPED)) {
 		zero_user_segment(page, cur, end);
+		++spiltted;
+		tight = false;
 		goto next_part;
 	}
 	if (map->m_flags & EROFS_MAP_FRAGMENT) {
-- 
2.24.4

[PATCH 6.1.y] erofs: ensure that the post-EOF tails are all zeroed

[Gao Xiang]

commit e4c1cf523d820730a86cae2c6d55924833b6f7ac upstream.

This was accidentally fixed up in commit e4c1cf523d82 but we can't
take the full change due to other dependancy issues, so here is just
the actual bugfix that is needed.

[Background]

keltargw reported an issue [1] that with mmaped I/Os, sometimes the
tail of the last page (after file ends) is not filled with zeroes.

The root cause is that such tail page could be wrongly selected for
inplace I/Os so the zeroed part will then be filled with compressed
data instead of zeroes.

A simple fix is to avoid doing inplace I/Os for such tail parts,
actually that was already fixed upstream in commit e4c1cf523d82
("erofs: tidy up z_erofs_do_read_page()") by accident.

[1] https://lore.kernel.org/r/3ad8b469-25db-a297-21f9-75db2d6ad224@linux.alibaba.com

Reported-by: keltargw <keltar.gw@gmail.com>
Fixes: 3883a79abd02 ("staging: erofs: introduce VLE decompression support")
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
---
 fs/erofs/zdata.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/fs/erofs/zdata.c b/fs/erofs/zdata.c
index 361f3c29897e..1b91ac5be961 100644
--- a/fs/erofs/zdata.c
+++ b/fs/erofs/zdata.c
@@ -869,6 +869,8 @@ static int z_erofs_do_read_page(struct z_erofs_decompress_frontend *fe,
 	cur = end - min_t(erofs_off_t, offset + end - map->m_la, end);
 	if (!(map->m_flags & EROFS_MAP_MAPPED)) {
 		zero_user_segment(page, cur, end);
+		++spiltted;
+		tight = false;
 		goto next_part;
 	}
 	if (map->m_flags & EROFS_MAP_FRAGMENT) {
-- 
2.24.4

[PATCH 5.15.y] erofs: ensure that the post-EOF tails are all zeroed

[Gao Xiang]

commit e4c1cf523d820730a86cae2c6d55924833b6f7ac upstream.

This was accidentally fixed up in commit e4c1cf523d82 but we can't
take the full change due to other dependancy issues, so here is just
the actual bugfix that is needed.

[Background]

keltargw reported an issue [1] that with mmaped I/Os, sometimes the
tail of the last page (after file ends) is not filled with zeroes.

The root cause is that such tail page could be wrongly selected for
inplace I/Os so the zeroed part will then be filled with compressed
data instead of zeroes.

A simple fix is to avoid doing inplace I/Os for such tail parts,
actually that was already fixed upstream in commit e4c1cf523d82
("erofs: tidy up z_erofs_do_read_page()") by accident.

[1] https://lore.kernel.org/r/3ad8b469-25db-a297-21f9-75db2d6ad224@linux.alibaba.com

Reported-by: keltargw <keltar.gw@gmail.com>
Fixes: 3883a79abd02 ("staging: erofs: introduce VLE decompression support")
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
---
 fs/erofs/zdata.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/fs/erofs/zdata.c b/fs/erofs/zdata.c
index c247b1bf57cc..f6536b224586 100644
--- a/fs/erofs/zdata.c
+++ b/fs/erofs/zdata.c
@@ -716,6 +716,8 @@ static int z_erofs_do_read_page(struct z_erofs_decompress_frontend *fe,
 	cur = end - min_t(erofs_off_t, offset + end - map->m_la, end);
 	if (!(map->m_flags & EROFS_MAP_MAPPED)) {
 		zero_user_segment(page, cur, end);
+		++spiltted;
+		tight = false;
 		goto next_part;
 	}
 
-- 
2.24.4

[PATCH 5.10.y] erofs: ensure that the post-EOF tails are all zeroed

[Gao Xiang]

commit e4c1cf523d820730a86cae2c6d55924833b6f7ac upstream.

This was accidentally fixed up in commit e4c1cf523d82 but we can't
take the full change due to other dependancy issues, so here is just
the actual bugfix that is needed.

[Background]

keltargw reported an issue [1] that with mmaped I/Os, sometimes the
tail of the last page (after file ends) is not filled with zeroes.

The root cause is that such tail page could be wrongly selected for
inplace I/Os so the zeroed part will then be filled with compressed
data instead of zeroes.

A simple fix is to avoid doing inplace I/Os for such tail parts,
actually that was already fixed upstream in commit e4c1cf523d82
("erofs: tidy up z_erofs_do_read_page()") by accident.

[1] https://lore.kernel.org/r/3ad8b469-25db-a297-21f9-75db2d6ad224@linux.alibaba.com

Reported-by: keltargw <keltar.gw@gmail.com>
Fixes: 3883a79abd02 ("staging: erofs: introduce VLE decompression support")
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
---
 fs/erofs/zdata.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/fs/erofs/zdata.c b/fs/erofs/zdata.c
index 9cff92738259..da133950d514 100644
--- a/fs/erofs/zdata.c
+++ b/fs/erofs/zdata.c
@@ -632,6 +632,8 @@ static int z_erofs_do_read_page(struct z_erofs_decompress_frontend *fe,
 	cur = end - min_t(erofs_off_t, offset + end - map->m_la, end);
 	if (!(map->m_flags & EROFS_MAP_MAPPED)) {
 		zero_user_segment(page, cur, end);
+		++spiltted;
+		tight = false;
 		goto next_part;
 	}
 
-- 
2.24.4

[PATCH 5.4.y] erofs: ensure that the post-EOF tails are all zeroed

[Gao Xiang]

commit e4c1cf523d820730a86cae2c6d55924833b6f7ac upstream.

This was accidentally fixed up in commit e4c1cf523d82 but we can't
take the full change due to other dependancy issues, so here is just
the actual bugfix that is needed.

[Background]

keltargw reported an issue [1] that with mmaped I/Os, sometimes the
tail of the last page (after file ends) is not filled with zeroes.

The root cause is that such tail page could be wrongly selected for
inplace I/Os so the zeroed part will then be filled with compressed
data instead of zeroes.

A simple fix is to avoid doing inplace I/Os for such tail parts,
actually that was already fixed upstream in commit e4c1cf523d82
("erofs: tidy up z_erofs_do_read_page()") by accident.

[1] https://lore.kernel.org/r/3ad8b469-25db-a297-21f9-75db2d6ad224@linux.alibaba.com

Reported-by: keltargw <keltar.gw@gmail.com>
Fixes: 3883a79abd02 ("staging: erofs: introduce VLE decompression support")
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
---
 fs/erofs/zdata.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/fs/erofs/zdata.c b/fs/erofs/zdata.c
index dcc377094f90..fb718c3e3ebd 100644
--- a/fs/erofs/zdata.c
+++ b/fs/erofs/zdata.c
@@ -639,6 +639,8 @@ static int z_erofs_do_read_page(struct z_erofs_decompress_frontend *fe,
 	cur = end - min_t(erofs_off_t, offset + end - map->m_la, end);
 	if (!(map->m_flags & EROFS_MAP_MAPPED)) {
 		zero_user_segment(page, cur, end);
+		++spiltted;
+		tight = false;
 		goto next_part;
 	}
 
-- 
2.24.4

[PATCH 4.19.y] erofs: ensure that the post-EOF tails are all zeroed

[Gao Xiang]

commit e4c1cf523d820730a86cae2c6d55924833b6f7ac upstream.

This was accidentally fixed up in commit e4c1cf523d82 but we can't
take the full change due to other dependancy issues, so here is just
the actual bugfix that is needed.

[Background]

keltargw reported an issue [1] that with mmaped I/Os, sometimes the
tail of the last page (after file ends) is not filled with zeroes.

The root cause is that such tail page could be wrongly selected for
inplace I/Os so the zeroed part will then be filled with compressed
data instead of zeroes.

A simple fix is to avoid doing inplace I/Os for such tail parts,
actually that was already fixed upstream in commit e4c1cf523d82
("erofs: tidy up z_erofs_do_read_page()") by accident.

[1] https://lore.kernel.org/r/3ad8b469-25db-a297-21f9-75db2d6ad224@linux.alibaba.com

Reported-by: keltargw <keltar.gw@gmail.com>
Fixes: 3883a79abd02 ("staging: erofs: introduce VLE decompression support")
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
---
 drivers/staging/erofs/unzip_vle.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/staging/erofs/unzip_vle.c b/drivers/staging/erofs/unzip_vle.c
index 83e4d9384bd2..7ccc4a18a900 100644
--- a/drivers/staging/erofs/unzip_vle.c
+++ b/drivers/staging/erofs/unzip_vle.c
@@ -675,6 +675,8 @@ static int z_erofs_do_read_page(struct z_erofs_vle_frontend *fe,
 	cur = end - min_t(unsigned, offset + end - map->m_la, end);
 	if (unlikely(!(map->m_flags & EROFS_MAP_MAPPED))) {
 		zero_user_segment(page, cur, end);
+		++spiltted;
+		tight = false;
 		goto next_part;
 	}
 
-- 
2.24.4

Re: [PATCH 6.5.y] erofs: ensure that the post-EOF tails are all zeroed

[Greg Kroah-Hartman]

On Thu, Aug 31, 2023 at 07:29:53PM +0800, Gao Xiang wrote:
> commit e4c1cf523d820730a86cae2c6d55924833b6f7ac upstream.
> 
> This was accidentally fixed up in commit e4c1cf523d82 but we can't
> take the full change due to other dependancy issues, so here is just
> the actual bugfix that is needed.
> 
> [Background]
> 
> keltargw reported an issue [1] that with mmaped I/Os, sometimes the
> tail of the last page (after file ends) is not filled with zeroes.
> 
> The root cause is that such tail page could be wrongly selected for
> inplace I/Os so the zeroed part will then be filled with compressed
> data instead of zeroes.
> 
> A simple fix is to avoid doing inplace I/Os for such tail parts,
> actually that was already fixed upstream in commit e4c1cf523d82
> ("erofs: tidy up z_erofs_do_read_page()") by accident.
> 
> [1] https://lore.kernel.org/r/3ad8b469-25db-a297-21f9-75db2d6ad224@linux.alibaba.com
> 
> Reported-by: keltargw <keltar.gw@gmail.com>
> Fixes: 3883a79abd02 ("staging: erofs: introduce VLE decompression support")
> Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
> ---
>  fs/erofs/zdata.c | 2 ++
>  1 file changed, 2 insertions(+)

All now queued up, thanks.

greg k-h