From: Jan Kara <jack@suse.cz>
To: "Aneesh Kumar K.V" <aneesh.kumar@linux.ibm.com>
Cc: "Michal Suchánek" <msuchanek@suse.de>,
jack@suse.de, linuxppc-dev@lists.ozlabs.org, mpe@ellerman.id.au,
linux-nvdimm@lists.01.org
Subject: Re: [RFC PATCH 1/2] libnvdimm: Add prctl control for disabling synchronous fault support.
Date: Fri, 29 May 2020 11:52:50 +0200 [thread overview]
Message-ID: <20200529095250.GP14550@quack2.suse.cz> (raw)
In-Reply-To: <6183cf4a-d134-99e5-936e-ef35f530c2ec@linux.ibm.com>
Hi!
On Fri 29-05-20 15:07:31, Aneesh Kumar K.V wrote:
> Thanks Michal. I also missed Jeff in this email thread.
And I think you'll also need some of the sched maintainers for the prctl
bits...
> On 5/29/20 3:03 PM, Michal Suchánek wrote:
> > Adding Jan
> >
> > On Fri, May 29, 2020 at 11:11:39AM +0530, Aneesh Kumar K.V wrote:
> > > With POWER10, architecture is adding new pmem flush and sync instructions.
> > > The kernel should prevent the usage of MAP_SYNC if applications are not using
> > > the new instructions on newer hardware.
> > >
> > > This patch adds a prctl option MAP_SYNC_ENABLE that can be used to enable
> > > the usage of MAP_SYNC. The kernel config option is added to allow the user
> > > to control whether MAP_SYNC should be enabled by default or not.
> > >
> > > Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
...
> > > diff --git a/kernel/fork.c b/kernel/fork.c
> > > index 8c700f881d92..d5a9a363e81e 100644
> > > --- a/kernel/fork.c
> > > +++ b/kernel/fork.c
> > > @@ -963,6 +963,12 @@ __cacheline_aligned_in_smp DEFINE_SPINLOCK(mmlist_lock);
> > > static unsigned long default_dump_filter = MMF_DUMP_FILTER_DEFAULT;
> > > +#ifdef CONFIG_ARCH_MAP_SYNC_DISABLE
> > > +unsigned long default_map_sync_mask = MMF_DISABLE_MAP_SYNC_MASK;
> > > +#else
> > > +unsigned long default_map_sync_mask = 0;
> > > +#endif
> > > +
I'm not sure CONFIG is really the right approach here. For a distro that would
basically mean to disable MAP_SYNC for all PPC kernels unless application
explicitly uses the right prctl. Shouldn't we rather initialize
default_map_sync_mask on boot based on whether the CPU we run on requires
new flush instructions or not? Otherwise the patch looks sensible.
Honza
--
Jan Kara <jack@suse.com>
SUSE Labs, CR
_______________________________________________
Linux-nvdimm mailing list -- linux-nvdimm@lists.01.org
To unsubscribe send an email to linux-nvdimm-leave@lists.01.org
WARNING: multiple messages have this Message-ID (diff)
From: Jan Kara <jack@suse.cz>
To: "Aneesh Kumar K.V" <aneesh.kumar@linux.ibm.com>
Cc: linux-nvdimm@lists.01.org, jack@suse.de,
"Jeff Moyer" <jmoyer@redhat.com>,
oohall@gmail.com, dan.j.williams@intel.com,
"Michal Suchánek" <msuchanek@suse.de>,
linuxppc-dev@lists.ozlabs.org
Subject: Re: [RFC PATCH 1/2] libnvdimm: Add prctl control for disabling synchronous fault support.
Date: Fri, 29 May 2020 11:52:50 +0200 [thread overview]
Message-ID: <20200529095250.GP14550@quack2.suse.cz> (raw)
In-Reply-To: <6183cf4a-d134-99e5-936e-ef35f530c2ec@linux.ibm.com>
Hi!
On Fri 29-05-20 15:07:31, Aneesh Kumar K.V wrote:
> Thanks Michal. I also missed Jeff in this email thread.
And I think you'll also need some of the sched maintainers for the prctl
bits...
> On 5/29/20 3:03 PM, Michal Suchánek wrote:
> > Adding Jan
> >
> > On Fri, May 29, 2020 at 11:11:39AM +0530, Aneesh Kumar K.V wrote:
> > > With POWER10, architecture is adding new pmem flush and sync instructions.
> > > The kernel should prevent the usage of MAP_SYNC if applications are not using
> > > the new instructions on newer hardware.
> > >
> > > This patch adds a prctl option MAP_SYNC_ENABLE that can be used to enable
> > > the usage of MAP_SYNC. The kernel config option is added to allow the user
> > > to control whether MAP_SYNC should be enabled by default or not.
> > >
> > > Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
...
> > > diff --git a/kernel/fork.c b/kernel/fork.c
> > > index 8c700f881d92..d5a9a363e81e 100644
> > > --- a/kernel/fork.c
> > > +++ b/kernel/fork.c
> > > @@ -963,6 +963,12 @@ __cacheline_aligned_in_smp DEFINE_SPINLOCK(mmlist_lock);
> > > static unsigned long default_dump_filter = MMF_DUMP_FILTER_DEFAULT;
> > > +#ifdef CONFIG_ARCH_MAP_SYNC_DISABLE
> > > +unsigned long default_map_sync_mask = MMF_DISABLE_MAP_SYNC_MASK;
> > > +#else
> > > +unsigned long default_map_sync_mask = 0;
> > > +#endif
> > > +
I'm not sure CONFIG is really the right approach here. For a distro that would
basically mean to disable MAP_SYNC for all PPC kernels unless application
explicitly uses the right prctl. Shouldn't we rather initialize
default_map_sync_mask on boot based on whether the CPU we run on requires
new flush instructions or not? Otherwise the patch looks sensible.
Honza
--
Jan Kara <jack@suse.com>
SUSE Labs, CR
next prev parent reply other threads:[~2020-05-29 9:52 UTC|newest]
Thread overview: 40+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-05-29 5:41 [RFC PATCH 1/2] libnvdimm: Add prctl control for disabling synchronous fault support Aneesh Kumar K.V
2020-05-29 5:41 ` Aneesh Kumar K.V
2020-05-29 5:41 ` [RFC PATCH 2/2] powerpc/pmem: Disable synchronous fault by default Aneesh Kumar K.V
2020-05-29 5:41 ` Aneesh Kumar K.V
2020-05-29 9:33 ` [RFC PATCH 1/2] libnvdimm: Add prctl control for disabling synchronous fault support Michal Suchánek
2020-05-29 9:33 ` Michal Suchánek
2020-05-29 9:37 ` Aneesh Kumar K.V
2020-05-29 9:37 ` Aneesh Kumar K.V
2020-05-29 9:52 ` Jan Kara [this message]
2020-05-29 9:52 ` Jan Kara
2020-05-29 10:55 ` Aneesh Kumar K.V
2020-05-29 10:55 ` Aneesh Kumar K.V
2020-05-29 19:22 ` Dan Williams
2020-05-29 19:22 ` Dan Williams
2020-05-30 7:18 ` Aneesh Kumar K.V
2020-05-30 7:18 ` Aneesh Kumar K.V
2020-05-30 16:35 ` Dan Williams
2020-05-30 16:35 ` Dan Williams
2020-06-01 9:50 ` Jan Kara
2020-06-01 9:50 ` Jan Kara
2020-06-02 17:59 ` Williams, Dan J
2020-06-02 17:59 ` Williams, Dan J
2020-06-03 8:26 ` Jan Kara
2020-06-03 8:26 ` Jan Kara
2020-06-03 9:09 ` Aneesh Kumar K.V
2020-06-03 9:09 ` Aneesh Kumar K.V
2020-06-08 7:42 ` Aneesh Kumar K.V
2020-06-08 7:42 ` Aneesh Kumar K.V
2020-06-01 10:09 ` Jan Kara
2020-06-01 10:09 ` Jan Kara
2020-06-01 12:01 ` Aneesh Kumar K.V
2020-06-01 12:01 ` Aneesh Kumar K.V
2020-06-01 12:07 ` Michal Suchánek
2020-06-01 12:07 ` Michal Suchánek
2020-06-01 12:20 ` Aneesh Kumar K.V
2020-06-01 12:20 ` Aneesh Kumar K.V
2020-06-02 7:57 ` Aneesh Kumar K.V
2020-06-02 7:57 ` Aneesh Kumar K.V
2020-06-01 14:56 ` Jan Kara
2020-06-01 14:56 ` Jan Kara
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20200529095250.GP14550@quack2.suse.cz \
--to=jack@suse.cz \
--cc=aneesh.kumar@linux.ibm.com \
--cc=jack@suse.de \
--cc=linux-nvdimm@lists.01.org \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=mpe@ellerman.id.au \
--cc=msuchanek@suse.de \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.