From: inki.dae@samsung•com (Inki Dae)
To: linux-arm-kernel@lists•infradead.org
Subject: [RFC 1/2] arm: cacheflush syscall: process only pages that are in the memory
Date: Wed, 31 Jan 2018 15:03:23 +0900 [thread overview]
Message-ID: <5A715C2B.70609@samsung.com> (raw)
In-Reply-To: <20180126213911.GY17719@n2100.armlinux.org.uk>
Hi Russell,
2018? 01? 27? 06:39? Russell King - ARM Linux ?(?) ? ?:
> On Fri, Jan 26, 2018 at 02:30:47PM +0100, Marek Szyprowski wrote:
>> Hi Russell,
>>
>> On 2018-01-26 12:32, Russell King - ARM Linux wrote:
>>> On Fri, Jan 26, 2018 at 12:14:40PM +0100, Marek Szyprowski wrote:
>>>> glibc in calls cacheflush syscall on the whole textrels section of the
>>>> relocated binaries. However, relocation usually doesn't touch all pages
>>>> of that section, so not all of them are read to memory when calling this
>>>> syscall. However flush_cache_user_range() function will unconditionally
>>>> touch all pages from the provided range, resulting additional overhead
>>>> related to reading all clean pages. Optimize this by calling
>>>> flush_cache_user_range() only on the pages that are already in the
>>>> memory.
>>> What ensures that another CPU doesn't remove a page while we're
>>> flushing it? That will trigger a data abort, which will want to
>>> take the mmap_sem, causing a deadlock.
>>
>> I thought that taking mmap_sem will prevent pages from being removed.
>> mmap_sem has been already taken in the previous implementation of that
>> syscall, until code simplification done by commit 97c72d89ce0e ("ARM:
>> cacheflush: don't bother rounding to nearest vma").
>
> No, you're not reading the previous code state correctly. Take a closer
> look at that commit.
>
> find_vma() requires that mmap_sem is held across the call as the VMA
> list is not stable without that semaphore held. However, more
> importantly, notice that it drops the semaphore _before_ calling the
> cache flushing function (__do_cache_op()).
>
> The point is that if __do_cache_op() faults, it will enter
> do_page_fault(), which will try to take the mmap_sem again, causing
> a deadlock.
I'm not sure but seems this patch tries to do cache-flush only in-memory pages.
So I think the page fault wouldn't happen becasue flush_cache_user_range function returns always 0.
Thanks,
Inki Dae
>
prev parent reply other threads:[~2018-01-31 6:03 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <CGME20180126111453eucas1p1330d8561386c3cf2bb457bc22c0d99a8@eucas1p1.samsung.com>
2018-01-26 11:14 ` [RFC 1/2] arm: cacheflush syscall: process only pages that are in the memory Marek Szyprowski
2018-01-26 11:14 ` [RFC 2/2] arm64: compat: " Marek Szyprowski
2018-01-26 17:41 ` Catalin Marinas
2018-01-26 18:02 ` Catalin Marinas
2018-01-26 11:32 ` [RFC 1/2] arm: " Russell King - ARM Linux
2018-01-26 13:30 ` Marek Szyprowski
2018-01-26 21:39 ` Russell King - ARM Linux
2018-01-31 6:03 ` Inki Dae [this message]
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=5A715C2B.70609@samsung.com \
--to=inki.dae@samsung$(echo .)com \
--cc=linux-arm-kernel@lists$(echo .)infradead.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox