From: Dan Williams <dan.j.williams@intel.com>
To: Peter Zijlstra <peterz@infradead.org>
Cc: Jan Kara <jack@suse.cz>, linux-nvdimm <linux-nvdimm@lists.01.org>,
david <david@fromorbit.com>,
Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
linux-xfs <linux-xfs@vger.kernel.org>,
Ingo Molnar <mingo@redhat.com>,
linux-fsdevel <linux-fsdevel@vger.kernel.org>,
Christoph Hellwig <hch@lst.de>
Subject: Re: [PATCH v5 08/11] wait_bit: introduce {wait_on, wake_up}_atomic_one
Date: Sun, 11 Mar 2018 10:15:55 -0700 [thread overview]
Message-ID: <CAPcyv4jr53dqVTzdEQ6Ln1Qg7nWVMQSVTkf6uc+Lg6Y-xruhuw@mail.gmail.com> (raw)
In-Reply-To: <20180311112725.GC4043@hirez.programming.kicks-ass.net>
On Sun, Mar 11, 2018 at 4:27 AM, Peter Zijlstra <peterz@infradead.org> wrote:
> On Fri, Mar 09, 2018 at 10:55:32PM -0800, Dan Williams wrote:
>> Add a generic facility for awaiting an atomic_t to reach a value of 1.
>>
>> Page reference counts typically need to reach 0 to be considered a
>> free / inactive page. However, ZONE_DEVICE pages allocated via
>> devm_memremap_pages() are never 'onlined', i.e. the put_page() typically
>> done at init time to assign pages to the page allocator is skipped.
>>
>> These pages will have their reference count elevated > 1 by
>> get_user_pages() when they are under DMA. In order to coordinate DMA to
>> these pages vs filesytem operations like hole-punch and truncate the
>> filesystem-dax implementation needs to capture the DMA-idle event i.e.
>> the 2 to 1 count transition).
>>
>> For now, this implementation does not have functional behavior change,
>> follow-on patches will add waiters for these page-idle events.
>
> Argh, no no no.. That whole wait_for_atomic_t thing is a giant
> trainwreck already and now you're making it worse still.
>
> Please have a look here:
>
> https://lkml.kernel.org/r/20171101190644.chwhfpoz3ywxx2m7@hirez.programming.kicks-ass.net
That thread seems to be worried about the object disappearing the
moment it's reference count reaches a target. That isn't the case with
the memmap / struct page objects for ZONE_DEVICE pages. I understand
wait_for_atomic_one() is broken in the general case, but as far as I
can see it works fine specifically for ZONE_DEVICE page busy tracking,
just not generic object lifetime.
_______________________________________________
Linux-nvdimm mailing list
Linux-nvdimm@lists.01.org
https://lists.01.org/mailman/listinfo/linux-nvdimm
WARNING: multiple messages have this Message-ID (diff)
From: Dan Williams <dan.j.williams@intel.com>
To: Peter Zijlstra <peterz@infradead.org>
Cc: linux-nvdimm <linux-nvdimm@lists.01.org>,
Ingo Molnar <mingo@redhat.com>, Christoph Hellwig <hch@lst.de>,
david <david@fromorbit.com>,
linux-xfs <linux-xfs@vger.kernel.org>,
linux-fsdevel <linux-fsdevel@vger.kernel.org>,
Jan Kara <jack@suse.cz>,
Ross Zwisler <ross.zwisler@linux.intel.com>,
Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH v5 08/11] wait_bit: introduce {wait_on,wake_up}_atomic_one
Date: Sun, 11 Mar 2018 10:15:55 -0700 [thread overview]
Message-ID: <CAPcyv4jr53dqVTzdEQ6Ln1Qg7nWVMQSVTkf6uc+Lg6Y-xruhuw@mail.gmail.com> (raw)
In-Reply-To: <20180311112725.GC4043@hirez.programming.kicks-ass.net>
On Sun, Mar 11, 2018 at 4:27 AM, Peter Zijlstra <peterz@infradead.org> wrote:
> On Fri, Mar 09, 2018 at 10:55:32PM -0800, Dan Williams wrote:
>> Add a generic facility for awaiting an atomic_t to reach a value of 1.
>>
>> Page reference counts typically need to reach 0 to be considered a
>> free / inactive page. However, ZONE_DEVICE pages allocated via
>> devm_memremap_pages() are never 'onlined', i.e. the put_page() typically
>> done at init time to assign pages to the page allocator is skipped.
>>
>> These pages will have their reference count elevated > 1 by
>> get_user_pages() when they are under DMA. In order to coordinate DMA to
>> these pages vs filesytem operations like hole-punch and truncate the
>> filesystem-dax implementation needs to capture the DMA-idle event i.e.
>> the 2 to 1 count transition).
>>
>> For now, this implementation does not have functional behavior change,
>> follow-on patches will add waiters for these page-idle events.
>
> Argh, no no no.. That whole wait_for_atomic_t thing is a giant
> trainwreck already and now you're making it worse still.
>
> Please have a look here:
>
> https://lkml.kernel.org/r/20171101190644.chwhfpoz3ywxx2m7@hirez.programming.kicks-ass.net
That thread seems to be worried about the object disappearing the
moment it's reference count reaches a target. That isn't the case with
the memmap / struct page objects for ZONE_DEVICE pages. I understand
wait_for_atomic_one() is broken in the general case, but as far as I
can see it works fine specifically for ZONE_DEVICE page busy tracking,
just not generic object lifetime.
next prev parent reply other threads:[~2018-03-11 17:09 UTC|newest]
Thread overview: 80+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-03-10 6:54 [PATCH v5 00/11] dax: fix dma vs truncate/hole-punch Dan Williams
2018-03-10 6:54 ` Dan Williams
2018-03-10 6:54 ` Dan Williams
2018-03-10 6:54 ` Dan Williams
2018-03-10 6:54 ` [PATCH v5 01/11] dax: store pfns in the radix Dan Williams
2018-03-10 6:54 ` Dan Williams
2018-03-10 6:54 ` Dan Williams
2018-03-10 6:54 ` [PATCH v5 02/11] xfs, dax: introduce xfs_dax_aops Dan Williams
2018-03-10 6:54 ` Dan Williams
2018-03-10 6:54 ` Dan Williams
2018-03-10 9:46 ` Christoph Hellwig
2018-03-10 9:46 ` Christoph Hellwig
2018-03-10 17:40 ` Dan Williams
2018-03-10 17:40 ` Dan Williams
2018-03-11 19:16 ` Dan Williams
2018-03-11 19:16 ` Dan Williams
2018-03-12 7:51 ` Christoph Hellwig
2018-03-12 7:51 ` Christoph Hellwig
2018-03-10 6:55 ` [PATCH v5 03/11] ext4, dax: introduce ext4_dax_aops Dan Williams
2018-03-10 6:55 ` Dan Williams
2018-03-10 6:55 ` Dan Williams
2018-03-10 6:55 ` Dan Williams
2018-03-10 6:55 ` [PATCH v5 04/11] ext2, dax: introduce ext2_dax_aops Dan Williams
2018-03-10 6:55 ` Dan Williams
2018-03-10 6:55 ` Dan Williams
2018-03-10 6:55 ` [PATCH v5 05/11] fs, dax: use page->mapping to warn if truncate collides with a busy page Dan Williams
2018-03-10 6:55 ` Dan Williams
2018-03-10 6:55 ` Dan Williams
2018-03-10 6:55 ` [PATCH v5 06/11] mm, dax: enable filesystems to trigger dev_pagemap ->page_free callbacks Dan Williams
2018-03-10 6:55 ` Dan Williams
2018-03-10 6:55 ` Dan Williams
2018-03-12 14:09 ` Jerome Glisse
2018-03-12 14:09 ` Jerome Glisse
2018-03-12 14:09 ` Jerome Glisse
2018-03-10 6:55 ` [PATCH v5 07/11] mm, dev_pagemap: introduce CONFIG_DEV_PAGEMAP_OPS Dan Williams
2018-03-10 6:55 ` Dan Williams
2018-03-10 6:55 ` Dan Williams
2018-03-12 14:17 ` Jerome Glisse
2018-03-12 14:17 ` Jerome Glisse
2018-03-12 14:17 ` Jerome Glisse
2018-03-12 18:17 ` Dan Williams
2018-03-12 18:17 ` Dan Williams
2018-03-10 6:55 ` [PATCH v5 08/11] wait_bit: introduce {wait_on,wake_up}_atomic_one Dan Williams
2018-03-10 6:55 ` Dan Williams
2018-03-10 6:55 ` Dan Williams
2018-03-11 11:27 ` [PATCH v5 08/11] wait_bit: introduce {wait_on, wake_up}_atomic_one Peter Zijlstra
2018-03-11 11:27 ` [PATCH v5 08/11] wait_bit: introduce {wait_on,wake_up}_atomic_one Peter Zijlstra
2018-03-11 17:15 ` Dan Williams [this message]
2018-03-11 17:15 ` Dan Williams
2018-03-12 19:32 ` [PATCH v5 08/11] wait_bit: introduce {wait_on, wake_up}_atomic_one Dan Williams
2018-03-12 19:32 ` [PATCH v5 08/11] wait_bit: introduce {wait_on,wake_up}_atomic_one Dan Williams
2018-03-13 10:20 ` [RFC][PATCH] sched/wait_bit: Introduce wait_var_event()/wake_up_var() Peter Zijlstra
2018-03-13 10:20 ` Peter Zijlstra
2018-03-14 4:12 ` Dan Williams
2018-03-14 4:12 ` Dan Williams
2018-03-15 5:46 ` Dan Williams
2018-03-15 5:46 ` Dan Williams
2018-03-15 9:58 ` David Howells
2018-03-15 9:58 ` David Howells
2018-03-15 11:19 ` Peter Zijlstra
2018-03-15 11:19 ` Peter Zijlstra
2018-03-15 11:51 ` Peter Zijlstra
2018-03-15 11:51 ` Peter Zijlstra
2018-03-15 14:45 ` David Howells
2018-03-15 14:45 ` David Howells
2018-03-15 14:53 ` Peter Zijlstra
2018-03-15 14:53 ` Peter Zijlstra
2018-03-10 6:55 ` [PATCH v5 09/11] mm, fs, dax: handle layout changes to pinned dax mappings Dan Williams
2018-03-10 6:55 ` Dan Williams
2018-03-10 6:55 ` Dan Williams
2018-03-10 6:55 ` [PATCH v5 10/11] xfs: prepare xfs_break_layouts() for another layout type Dan Williams
2018-03-10 6:55 ` Dan Williams
2018-03-10 6:55 ` Dan Williams
2018-03-10 9:51 ` Christoph Hellwig
2018-03-10 9:51 ` Christoph Hellwig
2018-03-10 6:55 ` [PATCH v5 11/11] xfs, dax: introduce xfs_break_dax_layouts() Dan Williams
2018-03-10 6:55 ` Dan Williams
2018-03-10 6:55 ` Dan Williams
2018-03-10 9:55 ` Christoph Hellwig
2018-03-10 9:55 ` Christoph Hellwig
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=CAPcyv4jr53dqVTzdEQ6Ln1Qg7nWVMQSVTkf6uc+Lg6Y-xruhuw@mail.gmail.com \
--to=dan.j.williams@intel.com \
--cc=david@fromorbit.com \
--cc=hch@lst.de \
--cc=jack@suse.cz \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-nvdimm@lists.01.org \
--cc=linux-xfs@vger.kernel.org \
--cc=mingo@redhat.com \
--cc=peterz@infradead.org \
/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.