Before this patch, asserts based on glocks did not print the glock with
the error. This patch introduces a new macro, gfs2_glock_assert_withdraw
which first prints the glock, then takes the assert.
This also changes a few glock asserts to the new macro.
Signed-off-by: Bob Peterson
---
fs/gfs2/g
This patch adds a new file: /sys/fs/gfs2/*/status which will report
the status of the file system. Catting this file dumps the current
status of the file system according to various superblock variables.
For example:
Journal Checked: 1
Journal Live: 1
Withdrawn: 0
No barriers:
Before this patch, transactions could be merged into the system
transaction by function gfs2_merge_trans(), but then it is not
considered attached. The caller, log_refund, only sets TR_ATTACHED
if the transaction is not merged. Later, in function, gfs2_trans_end,
if TR_ATTACHED is not set (because
Hi Andreas,
- Original Message -
(snip)
> > @@ -970,7 +969,16 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct
> > gfs2_glock *gl, u32 flags)
> >
> > if (!(flags & GFS2_LOG_HEAD_FLUSH_NORMAL)) {
> > if (!sdp->sd_log_idle) {
> > + unsigned lo
On Fri, Jun 5, 2020 at 4:49 PM Bob Peterson wrote:
> Hi Andreas,
>
> - Original Message -
> (snip)
> > > @@ -970,7 +969,16 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct
> > > gfs2_glock *gl, u32 flags)
> > >
> > > if (!(flags & GFS2_LOG_HEAD_FLUSH_NORMAL)) {
> > >
This patch adds a new slab for gfs2 transactions. That allows us to
have an initialization function and protect against some errors.
Signed-off-by: Bob Peterson
---
fs/gfs2/glops.c | 2 ++
fs/gfs2/log.c | 9 +
fs/gfs2/main.c | 9 +
fs/gfs2/trans.c | 22 ++---
I've revised my recent patch set based on feedback from Andreas.
Here is my latest. Changes from last time are:
1. Removed the spin_trylock related info from the status sysfs file
as per Andreas's feedback. Although valuable to have, I see his point.
2. Removed "gfs2: Add new trace point for gl
Bob,
this very much looks like a debugging trace point that has no business
in production code.
Thanks,
Andreas
Hi,
I revised the patch description for Andreas:
This patch adds a new slab for gfs2 transactions. That allows us to
reduce kernel memory fragmentation, have better organization of data
for analysis of vmcore dumps. A new centralized function is added to
free the slab objects, and it exposes use-
- Original Message -
> On Fri, Jun 5, 2020 at 4:49 PM Bob Peterson wrote:
> > Hi Andreas,
> >
> > - Original Message -
> > (snip)
> > > > @@ -970,7 +969,16 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct
> > > > gfs2_glock *gl, u32 flags)
> > > >
> > > > if (!(flags & G
On Fri, Jun 5, 2020 at 6:15 PM Bob Peterson wrote:
> - Original Message -
> > On Fri, Jun 5, 2020 at 4:49 PM Bob Peterson wrote:
> > > Hi Andreas,
> > >
> > > - Original Message -
> > > (snip)
> > > > > @@ -970,7 +969,16 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct
> > > >
Hi Bob,
On Tue, May 26, 2020 at 3:05 PM Bob Peterson wrote:
> This adds checks for gfs2_log_flush being stuck, similarly to the check
> in gfs2_ail1_flush.
>
> Signed-off-by: Bob Peterson
> ---
> fs/gfs2/log.c | 14 +++---
> 1 file changed, 11 insertions(+), 3 deletions(-)
>
> diff --gi
Hi Bob,
On Tue, May 26, 2020 at 3:07 PM Bob Peterson wrote:
> Andreas expressed some concerns about some of the others. For example, he
> didn't like that the new "status" sysfs file was taking "try" locks, but
> if the lock is held, I don't know of a better way to do this.
walking a linked list
This patch makes the glock dumps in debugfs print the number of pages
(nrpages) for address space glocks. This will aid in debugging.
Signed-off-by: Bob Peterson
---
fs/gfs2/glock.c | 25 -
1 file changed, 16 insertions(+), 9 deletions(-)
diff --git a/fs/gfs2/glock.c b/f
This adds checks for gfs2_log_flush being stuck, similarly to the check
in gfs2_ail1_flush. To faciliate this and make the strings easy to grep
we move the ail1 emptying to its own function, empty_ail1_list.
Signed-off-by: Bob Peterson
---
fs/gfs2/log.c | 34 +-
1
Hi,
I've posted similar patches for this in the past, but they all had
shortcomings. This is my latest version. There is a follow-on patch
which I will post shortly. In the future, I want to add a new
slab initialization function, but for now this is enough.
Note that the use-after-free warnings
From: Andreas Gruenbacher
[ Upstream commit 20be493b787cd581c9fffad7fcd6bfbe6af1050c ]
Fix several issues in the previous gfs2_find_jhead fix:
* When updating @blocks_submitted, @block refers to the first block block not
submitted yet, not the last block submitted, so fix an off-by-one error.
From: Andreas Gruenbacher
[ Upstream commit 20be493b787cd581c9fffad7fcd6bfbe6af1050c ]
Fix several issues in the previous gfs2_find_jhead fix:
* When updating @blocks_submitted, @block refers to the first block block not
submitted yet, not the last block submitted, so fix an off-by-one error.
Hi,
Before this patch, transactions could be merged into the system
transaction by function gfs2_merge_trans(), but then it is not
considered attached. The caller, log_refund, only sets TR_ATTACHED
if the transaction is not merged. Later, in function, gfs2_trans_end,
if TR_ATTACHED is not set (bec
19 matches
Mail list logo
