* Re: [PATCH] xfs_db: don't hardcode 'type data' size at 512b
2024-01-23 4:10 ` [PATCH] xfs_db: don't hardcode 'type data' size at 512b Darrick J. Wong
2024-01-23 4:15 ` Christoph Hellwig
@ 2024-01-23 10:38 ` Carlos Maiolino
2024-02-22 22:04 ` [PATCH v2] " Darrick J. Wong
2 siblings, 0 replies; 4+ messages in thread
From: Carlos Maiolino @ 2024-01-23 10:38 UTC (permalink / raw)
To: Darrick J. Wong; +Cc: Christoph Hellwig, linux-xfs
On Mon, Jan 22, 2024 at 08:10:44PM -0800, Darrick J. Wong wrote:
> From: Darrick J. Wong <djwong@kernel.org>
>
> On a disk with 4096-byte LBAs, the xfs_db 'type data' subcommand doesn't
> work:
>
> # xfs_io -c 'sb' -c 'type data' /dev/sda
> xfs_db: read failed: Invalid argument
> no current object
>
> The cause of this is the hardcoded initialization of bb_count when we're
> setting type data -- it should be the filesystem sector size, not just 1.
>
> Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Looks good:
Reviewed-by: Carlos Maiolino <cmaiolino@redhat.com>
> ---
> db/block.c | 7 ++++---
> db/io.c | 3 ++-
> 2 files changed, 6 insertions(+), 4 deletions(-)
>
> diff --git a/db/block.c b/db/block.c
> index f234fcb4edcb..c847a91dd22f 100644
> --- a/db/block.c
> +++ b/db/block.c
> @@ -149,6 +149,7 @@ daddr_f(
> int64_t d;
> char *p;
> int c;
> + int bb_count = BTOBB(mp->m_sb.sb_sectsize);
> xfs_rfsblock_t max_daddrs = mp->m_sb.sb_dblocks;
> enum daddr_target tgt = DT_DATA;
>
> @@ -202,13 +203,13 @@ daddr_f(
> ASSERT(typtab[TYP_DATA].typnm == TYP_DATA);
> switch (tgt) {
> case DT_DATA:
> - set_cur(&typtab[TYP_DATA], d, 1, DB_RING_ADD, NULL);
> + set_cur(&typtab[TYP_DATA], d, bb_count, DB_RING_ADD, NULL);
> break;
> case DT_RT:
> - set_rt_cur(&typtab[TYP_DATA], d, 1, DB_RING_ADD, NULL);
> + set_rt_cur(&typtab[TYP_DATA], d, bb_count, DB_RING_ADD, NULL);
> break;
> case DT_LOG:
> - set_log_cur(&typtab[TYP_DATA], d, 1, DB_RING_ADD, NULL);
> + set_log_cur(&typtab[TYP_DATA], d, bb_count, DB_RING_ADD, NULL);
> break;
> }
> return 0;
> diff --git a/db/io.c b/db/io.c
> index 580d34015868..3841c0dcb86e 100644
> --- a/db/io.c
> +++ b/db/io.c
> @@ -681,7 +681,8 @@ void
> set_iocur_type(
> const typ_t *type)
> {
> - int bb_count = 1; /* type's size in basic blocks */
> + /* type's size in basic blocks */
> + int bb_count = BTOBB(mp->m_sb.sb_sectsize);
> int boff = iocur_top->boff;
>
> /*
>
^ permalink raw reply [flat|nested] 4+ messages in thread* [PATCH v2] xfs_db: don't hardcode 'type data' size at 512b
2024-01-23 4:10 ` [PATCH] xfs_db: don't hardcode 'type data' size at 512b Darrick J. Wong
2024-01-23 4:15 ` Christoph Hellwig
2024-01-23 10:38 ` Carlos Maiolino
@ 2024-02-22 22:04 ` Darrick J. Wong
2 siblings, 0 replies; 4+ messages in thread
From: Darrick J. Wong @ 2024-02-22 22:04 UTC (permalink / raw)
To: Christoph Hellwig; +Cc: cem, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
On a disk with 4096-byte LBAs, the xfs_db 'type data' subcommand doesn't
work:
# xfs_io -c 'sb' -c 'type data' /dev/sda
xfs_db: read failed: Invalid argument
no current object
The cause of this is the hardcoded initialization of bb_count when we're
setting type data -- it should be the filesystem sector size, not just 1.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Carlos Maiolino <cmaiolino@redhat.com>
---
v2: rebase to for-next
---
db/block.c | 3 ++-
db/io.c | 3 ++-
2 files changed, 4 insertions(+), 2 deletions(-)
diff --git a/db/block.c b/db/block.c
index d730c7796710..22930e5a287e 100644
--- a/db/block.c
+++ b/db/block.c
@@ -124,6 +124,7 @@ daddr_f(
{
int64_t d;
char *p;
+ int bb_count = BTOBB(mp->m_sb.sb_sectsize);
if (argc == 1) {
xfs_daddr_t daddr = iocur_top->off >> BBSHIFT;
@@ -144,7 +145,7 @@ daddr_f(
return 0;
}
ASSERT(typtab[TYP_DATA].typnm == TYP_DATA);
- set_cur(&typtab[TYP_DATA], d, 1, DB_RING_ADD, NULL);
+ set_cur(&typtab[TYP_DATA], d, bb_count, DB_RING_ADD, NULL);
return 0;
}
diff --git a/db/io.c b/db/io.c
index 590dd1f82f7b..9b2c6b4cf7e9 100644
--- a/db/io.c
+++ b/db/io.c
@@ -652,7 +652,8 @@ void
set_iocur_type(
const typ_t *type)
{
- int bb_count = 1; /* type's size in basic blocks */
+ /* type's size in basic blocks */
+ int bb_count = BTOBB(mp->m_sb.sb_sectsize);
int boff = iocur_top->boff;
/*
^ permalink raw reply related [flat|nested] 4+ messages in thread