[PATCH 5.4 0/1] mm: allow a controlled amount of unfairness in the page lock

Saeed Mirzamohammadi posted 1 patch 2 years, 3 months ago
There is a newer version of this series
include/linux/mm.h   |   2 +
include/linux/wait.h |   2 +
kernel/sysctl.c      |   8 +++
mm/filemap.c         | 160 ++++++++++++++++++++++++++++++++++---------
4 files changed, 141 insertions(+), 31 deletions(-)
[PATCH 5.4 0/1] mm: allow a controlled amount of unfairness in the page lock
Posted by Saeed Mirzamohammadi 2 years, 3 months ago
We observed a 35% of regression running phoronix pts/ramspeed and also 16%
with unixbench. Regression is caused by the following commit:
dd0f194cfeb5 | mm: rewrite wait_on_page_bit_common() logic

Backporting this fixes the regression (this is already in 5.9+):
- 5ef64cc8987a mm: allow a controlled amount of unfairness in the page lock

Linus Torvalds (1):
  mm: allow a controlled amount of unfairness in the page lock

 include/linux/mm.h   |   2 +
 include/linux/wait.h |   2 +
 kernel/sysctl.c      |   8 +++
 mm/filemap.c         | 160 ++++++++++++++++++++++++++++++++++---------
 4 files changed, 141 insertions(+), 31 deletions(-)

-- 
2.41.0
Re: [PATCH 5.4 0/1] mm: allow a controlled amount of unfairness in the page lock
Posted by Greg KH 2 years, 3 months ago
On Mon, Aug 21, 2023 at 03:25:45PM -0700, Saeed Mirzamohammadi wrote:
> We observed a 35% of regression running phoronix pts/ramspeed and also 16%
> with unixbench. Regression is caused by the following commit:
> dd0f194cfeb5 | mm: rewrite wait_on_page_bit_common() logic

That is not a valid git id in Linus's or in the linux-stable repo that I
can see.  Are you sure that it is correct?

thanks,

greg k-h
Re: [PATCH 5.4 0/1] mm: allow a controlled amount of unfairness in the page lock
Posted by Saeed Mirzamohammadi 2 years, 3 months ago

> On Aug 22, 2023, at 12:08 AM, Greg KH <gregkh@linuxfoundation.org> wrote:
> 
> On Mon, Aug 21, 2023 at 03:25:45PM -0700, Saeed Mirzamohammadi wrote:
>> We observed a 35% of regression running phoronix pts/ramspeed and also 16%
>> with unixbench. Regression is caused by the following commit:
>> dd0f194cfeb5 | mm: rewrite wait_on_page_bit_common() logic
> 
> That is not a valid git id in Linus's or in the linux-stable repo that I
> can see.  Are you sure that it is correct?

Sorry for the incorrect sha. Here are the correct ones:

  kernel_dot_org/linux-stable.git    linux-5.4.y            - c32ab1c1959a
  kernel_dot_org/torvalds_linux.git  master                 - 2a9127fcf229
  kernel_dot_org/linux-stable.git    master                 - 2a9127fcf229
---
  subject          : mm: rewrite wait_on_page_bit_common() logic
  author           : torvalds@linux-foundation.org
  author date      : 2020-07-23 17:16:49

Thanks,
Saeed

> 
> thanks,
> 
> greg k-h
Re: [PATCH 5.4 0/1] mm: allow a controlled amount of unfairness in the page lock
Posted by Greg KH 2 years, 3 months ago
On Tue, Aug 22, 2023 at 05:20:50PM +0000, Saeed Mirzamohammadi wrote:
> 
> 
> > On Aug 22, 2023, at 12:08 AM, Greg KH <gregkh@linuxfoundation.org> wrote:
> > 
> > On Mon, Aug 21, 2023 at 03:25:45PM -0700, Saeed Mirzamohammadi wrote:
> >> We observed a 35% of regression running phoronix pts/ramspeed and also 16%
> >> with unixbench. Regression is caused by the following commit:
> >> dd0f194cfeb5 | mm: rewrite wait_on_page_bit_common() logic
> > 
> > That is not a valid git id in Linus's or in the linux-stable repo that I
> > can see.  Are you sure that it is correct?
> 
> Sorry for the incorrect sha. Here are the correct ones:
> 
>   kernel_dot_org/linux-stable.git    linux-5.4.y            - c32ab1c1959a
>   kernel_dot_org/torvalds_linux.git  master                 - 2a9127fcf229
>   kernel_dot_org/linux-stable.git    master                 - 2a9127fcf229
> ---
>   subject          : mm: rewrite wait_on_page_bit_common() logic
>   author           : torvalds@linux-foundation.org
>   author date      : 2020-07-23 17:16:49

Thanks, now queued up.

greg k-h