[PATCH 2/2] zram: fixup read_block_state()

[PATCH 2/2] zram: fixup read_block_state()

[Sergey Senozhatsky]

ac_time is now in seconds, do not use ktime_to_timespec64()

Reported-by: Chris Mason <clm@meta.com>
Signed-off-by: Sergey Senozhatsky <senozhatsky@chromium.org>
---
 drivers/block/zram/zram_drv.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/drivers/block/zram/zram_drv.c b/drivers/block/zram/zram_drv.c
index cc51aa8b6181..912711faa4e4 100644
--- a/drivers/block/zram/zram_drv.c
+++ b/drivers/block/zram/zram_drv.c
@@ -1579,11 +1579,9 @@ static ssize_t read_block_state(struct file *file, char __user *buf,
 		if (!slot_allocated(zram, index))
 			goto next;
 
-		ts = ktime_to_timespec64(zram->table[index].attr.ac_time);
 		copied = snprintf(kbuf + written, count,
-			"%12zd %12lld.%06lu %c%c%c%c%c%c\n",
-			index, (s64)ts.tv_sec,
-			ts.tv_nsec / NSEC_PER_USEC,
+			"%12zd %12u.%06lu %c%c%c%c%c%c\n",
+			index, zram->table[index].attr.ac_time, 0,
 			test_slot_flag(zram, index, ZRAM_SAME) ? 's' : '.',
 			test_slot_flag(zram, index, ZRAM_WB) ? 'w' : '.',
 			test_slot_flag(zram, index, ZRAM_HUGE) ? 'h' : '.',
-- 
2.52.0.457.g6b5491de43-goog

Re: [PATCH 2/2] zram: fixup read_block_state()

[Andrew Morton]

On Thu, 15 Jan 2026 12:30:06 +0900 Sergey Senozhatsky <senozhatsky@chromium.org> wrote:

> ac_time is now in seconds, do not use ktime_to_timespec64()

Please help me out here - which patch introduced the issue which Chris
found?

> Reported-by: Chris Mason <clm@meta.com>

The Closes: link would be helpful.

> Signed-off-by: Sergey Senozhatsky <senozhatsky@chromium.org>
> ---
>  drivers/block/zram/zram_drv.c | 6 ++----
>  1 file changed, 2 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/block/zram/zram_drv.c b/drivers/block/zram/zram_drv.c
> index cc51aa8b6181..912711faa4e4 100644
> --- a/drivers/block/zram/zram_drv.c
> +++ b/drivers/block/zram/zram_drv.c
> @@ -1579,11 +1579,9 @@ static ssize_t read_block_state(struct file *file, char __user *buf,
>  		if (!slot_allocated(zram, index))
>  			goto next;
>  
> -		ts = ktime_to_timespec64(zram->table[index].attr.ac_time);
>  		copied = snprintf(kbuf + written, count,
> -			"%12zd %12lld.%06lu %c%c%c%c%c%c\n",
> -			index, (s64)ts.tv_sec,
> -			ts.tv_nsec / NSEC_PER_USEC,
> +			"%12zd %12u.%06lu %c%c%c%c%c%c\n",
> +			index, zram->table[index].attr.ac_time, 0,

local `ts' is now unused?

>  			test_slot_flag(zram, index, ZRAM_SAME) ? 's' : '.',
>  			test_slot_flag(zram, index, ZRAM_WB) ? 'w' : '.',
>  			test_slot_flag(zram, index, ZRAM_HUGE) ? 'h' : '.',

Re: [PATCH 2/2] zram: fixup read_block_state()

[Sergey Senozhatsky]

On (26/01/15 14:07), Andrew Morton wrote:
> On Thu, 15 Jan 2026 12:30:06 +0900 Sergey Senozhatsky <senozhatsky@chromium.org> wrote:
> 
> > ac_time is now in seconds, do not use ktime_to_timespec64()
> 
> Please help me out here - which patch introduced the issue which Chris
> found?

This one:
https://git.kernel.org/pub/scm/linux/kernel/git/akpm/25-new.git/tree/patches/zram-use-u32-for-entry-ac_time-tracking.patch

> > Reported-by: Chris Mason <clm@meta.com>
> 
> The Closes: link would be helpful.

I thought that for fixup patches that are still in your tree we
don't need Closes: tags.  If you prefer to have them, it should
be this email:
http://lore.kernel.org/linux-kernel/20260114124522.1326519-1-clm@meta.com

[..]
> > -		ts = ktime_to_timespec64(zram->table[index].attr.ac_time);
> >  		copied = snprintf(kbuf + written, count,
> > -			"%12zd %12lld.%06lu %c%c%c%c%c%c\n",
> > -			index, (s64)ts.tv_sec,
> > -			ts.tv_nsec / NSEC_PER_USEC,
> > +			"%12zd %12u.%06lu %c%c%c%c%c%c\n",
> > +			index, zram->table[index].attr.ac_time, 0,
> 
> local `ts' is now unused?

Oh, yes.  Somehow my compiler didn't notice that.

Re: [PATCH 2/2] zram: fixup read_block_state()

[Andrew Morton]

On Fri, 16 Jan 2026 07:54:25 +0900 Sergey Senozhatsky <senozhatsky@chromium.org> wrote:

> On (26/01/15 14:07), Andrew Morton wrote:
> > On Thu, 15 Jan 2026 12:30:06 +0900 Sergey Senozhatsky <senozhatsky@chromium.org> wrote:
> > 
> > > ac_time is now in seconds, do not use ktime_to_timespec64()
> > 
> > Please help me out here - which patch introduced the issue which Chris
> > found?
> 
> This one:
> https://git.kernel.org/pub/scm/linux/kernel/git/akpm/25-new.git/tree/patches/zram-use-u32-for-entry-ac_time-tracking.patch

I figured that out when I fouf Chris's email ;)

> > > Reported-by: Chris Mason <clm@meta.com>
> > 
> > The Closes: link would be helpful.
> 
> I thought that for fixup patches that are still in your tree we
> don't need Closes: tags.  If you prefer to have them, it should
> be this email:
> http://lore.kernel.org/linux-kernel/20260114124522.1326519-1-clm@meta.com

Yep, got it thanks.

Closes: tags are helpful when reviewing, to see all the info.  Fixes:
tags aren't useful because the hashes are unstable and because the fix
is destined to become part of the thing which it fixes anyway.