From: Andy Lutomirski <luto@amacapital.net>
To: Chris Wilson <chris@chris-wilson.co.uk>,
"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
Ross Zwisler <ross.zwisler@linux.intel.com>,
"H . Peter Anvin" <hpa@linux.intel.com>,
Andy Lutomirski <luto@amacapital.net>,
Borislav Petkov <bp@alien8.de>, Brian Gerst <brgerst@gmail.com>,
Denys Vlasenko <dvlasenk@redhat.com>,
"H. Peter Anvin" <hpa@zytor.com>,
Linus Torvalds <torvalds@linux-foundation.org>,
Thomas Gleixner <tglx@linutronix.de>,
Imre Deak <imre.deak@intel.com>,
Daniel Vetter <daniel.vetter@ffwll.ch>,
DRI <dri-devel@lists.freedesktop.org>
Subject: Re: [PATCH] x86: Add an explicit barrier() to clflushopt()
Date: Thu, 7 Jan 2016 09:55:51 -0800 [thread overview]
Message-ID: <CALCETrU8otVfBe0+N-E0sc=j2L79qh7gqte7QSVb_YzGbJYNCA@mail.gmail.com> (raw)
In-Reply-To: <20160107101652.GF652@nuc-i3427.alporthouse.com>
On Thu, Jan 7, 2016 at 2:16 AM, Chris Wilson <chris@chris-wilson.co.uk> wrote:
> On Mon, Oct 19, 2015 at 10:58:55AM +0100, Chris Wilson wrote:
>> During testing we observed that the last cacheline was not being flushed
>> from a
>>
>> mb()
>> for (addr = addr & -clflush_size; addr < end; addr += clflush_size)
>> clflushopt();
>> mb()
>>
>> loop (where the initial addr and end were not cacheline aligned).
>>
>> Changing the loop from addr < end to addr <= end, or replacing the
>> clflushopt() with clflush() both fixed the testcase. Hinting that GCC
>> was miscompling the assembly within the loop and specifically the
>> alternative within clflushopt() was confusing the loop optimizer.
>>
>> Adding a barrier() into clflushopt() is enough for GCC to dtrt, but
>> solving why GCC is not seeing the constraints from the alternative_io()
>> would be smarter...
>>
>> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=92501
>> Testcase: gem_tiled_partial_pwrite_pread/read
>> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
>> Cc: Ross Zwisler <ross.zwisler@linux.intel.com>
>> Cc: H. Peter Anvin <hpa@linux.intel.com>
>> Cc: Imre Deak <imre.deak@intel.com>
>> Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
>> Cc: dri-devel@lists.freedesktop.org
>> ---
>> arch/x86/include/asm/special_insns.h | 5 +++++
>> 1 file changed, 5 insertions(+)
>>
>> diff --git a/arch/x86/include/asm/special_insns.h b/arch/x86/include/asm/special_insns.h
>> index 2270e41b32fd..0c7aedbf8930 100644
>> --- a/arch/x86/include/asm/special_insns.h
>> +++ b/arch/x86/include/asm/special_insns.h
>> @@ -199,6 +199,11 @@ static inline void clflushopt(volatile void *__p)
>> ".byte 0x66; clflush %P0",
>> X86_FEATURE_CLFLUSHOPT,
>> "+m" (*(volatile char __force *)__p));
>> + /* GCC (4.9.1 and 5.2.1 at least) appears to be very confused when
>> + * meeting this alternative() and demonstrably miscompiles loops
>> + * iterating over clflushopts.
>> + */
>> + barrier();
>> }
>
> Or an alternative:
>
> +#define alternative_output(oldinstr, newinstr, feature, output) \
> + asm volatile (ALTERNATIVE(oldinstr, newinstr, feature) \
> + : output : "i" (0) : "memory")
>
> I would really appreciate some knowledgeable folks taking a look at the
> asm for clflushopt() as it still affects today's kernel and gcc.
>
> Fwiw, I have confirmed that arch/x86/mm/pageattr.c clflush_cache_range()
> is similarly affected.
Unless I'm mis-reading the asm, clflush_cache_range() is compiled
correctly for me. (I don't know what the %P is for in the asm, but
that shouldn't matter.) The ALTERNATIVE shouldn't even be visible to
the optimizer.
Can you attach a bad .s file and let us know what gcc version this is?
(You can usually do 'make foo/bar/baz.s' to get a .s file.) I'd also
be curious whether changing clflushopt to clwb works around the issue.
--Andy
next prev parent reply other threads:[~2016-01-07 17:56 UTC|newest]
Thread overview: 29+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-10-19 9:58 [PATCH] x86: Add an explicit barrier() to clflushopt() Chris Wilson
2015-10-19 10:16 ` Borislav Petkov
2015-10-19 11:05 ` Chris Wilson
2015-10-19 11:25 ` Borislav Petkov
2015-10-19 18:29 ` Ross Zwisler
2016-01-07 10:16 ` Chris Wilson
2016-01-07 17:55 ` Andy Lutomirski [this message]
2016-01-07 19:44 ` Chris Wilson
2016-01-07 21:05 ` H. Peter Anvin
2016-01-07 21:54 ` Chris Wilson
2016-01-07 22:29 ` H. Peter Anvin
2016-01-07 22:32 ` H. Peter Anvin
2016-01-09 5:55 ` H. Peter Anvin
2016-01-09 8:01 ` Chris Wilson
2016-01-09 22:36 ` Andy Lutomirski
2016-01-11 11:28 ` Chris Wilson
2016-01-11 20:11 ` Linus Torvalds
2016-01-11 21:05 ` Chris Wilson
2016-01-12 16:37 ` Chris Wilson
2016-01-12 17:05 ` Linus Torvalds
2016-01-12 21:13 ` Chris Wilson
2016-01-12 22:07 ` Linus Torvalds
2016-01-13 0:55 ` Chris Wilson
2016-01-13 2:06 ` Linus Torvalds
2016-01-13 2:42 ` Andy Lutomirski
2016-01-13 4:39 ` Linus Torvalds
2016-01-13 12:34 ` Chris Wilson
2016-01-13 18:45 ` Linus Torvalds
2016-01-12 17:17 ` H. Peter Anvin
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='CALCETrU8otVfBe0+N-E0sc=j2L79qh7gqte7QSVb_YzGbJYNCA@mail.gmail.com' \
--to=luto@amacapital.net \
--cc=bp@alien8.de \
--cc=brgerst@gmail.com \
--cc=chris@chris-wilson.co.uk \
--cc=daniel.vetter@ffwll.ch \
--cc=dri-devel@lists.freedesktop.org \
--cc=dvlasenk@redhat.com \
--cc=hpa@linux.intel.com \
--cc=hpa@zytor.com \
--cc=imre.deak@intel.com \
--cc=linux-kernel@vger.kernel.org \
--cc=ross.zwisler@linux.intel.com \
--cc=tglx@linutronix.de \
--cc=torvalds@linux-foundation.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).