Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1592618

Question Regarding ERMS memcpy

Show all headers | View raw

[Multipart message — attachments visible in raw view] - view raw

Hi,

I'm trying to chase down a performance issue with a driver I'm working
on that does a repeated memcpy_fromio of about 1KB from a PCI device. I
made a small change from a fixed size copy to a variable size only to be
surprised with a performance decrease of about 1/3.

I've looked through the code and discovered this is due to switching out
__builtin_memcpy with __memcpy when the length is not constant. This
makes sense and I've now been looking into the inner workings of
memcpy_64.S.

The CPU I'm testing on is a Sandy Bridge and according to /proc/cpuinfo,
it does _not_ have the erms bit and does have the rep_good bit. So I
expect to be using the "rep movsq" version of memcpy. However, I've done
some testing with my driver by hacking in specific memcpy
implementations and I've found the following results:

__builtin_memcpy w/const length: 85KB/s
memcpy_fromio: 26kB/s
__builtin_memcpy: 26kB/s
memcpy_movsq: 126kB/s
memcpy_erms: 26kB/s

Thus, based on these performance numbers, it almost seems like my
platform is using the erms version when it probably shouldn't be.

So my question is: how do I find out what version of memcpy my actual
machine is using and fix it if it is wrong?

I'm running with a 4.10.0 kernel and I've attached my cpuinfo.

Thanks for any insights,

Logan

Back to linux.kernel | Previous | Next — Next in thread | Find similar | Unroll thread

Thread

Question Regarding ERMS memcpy Logan Gunthorpe <logang@deltatee.com> - 2017-03-04 22:20 +0100
  Re: Question Regarding ERMS memcpy Borislav Petkov <bp@suse.de> - 2017-03-04 23:50 +0100
    Re: Question Regarding ERMS memcpy hpa@zytor.com - 2017-03-05 01:00 +0100
      Re: Question Regarding ERMS memcpy Borislav Petkov <bp@suse.de> - 2017-03-05 01:20 +0100
        Re: Question Regarding ERMS memcpy hpa@zytor.com - 2017-03-05 01:30 +0100
          Re: Question Regarding ERMS memcpy Borislav Petkov <bp@suse.de> - 2017-03-05 01:40 +0100
            Re: Question Regarding ERMS memcpy hpa@zytor.com - 2017-03-05 02:00 +0100
              Re: Question Regarding ERMS memcpy Borislav Petkov <bp@suse.de> - 2017-03-05 11:00 +0100
                Re: Question Regarding ERMS memcpy Borislav Petkov <bp@suse.de> - 2017-03-05 12:20 +0100
                Re: Question Regarding ERMS memcpy Borislav Petkov <bp@suse.de> - 2017-03-05 14:10 +0100
                Re: Question Regarding ERMS memcpy Linus Torvalds <torvalds@linux-foundation.org> - 2017-03-05 20:30 +0100
                Re: Question Regarding ERMS memcpy Borislav Petkov <bp@suse.de> - 2017-03-05 21:20 +0100
                Re: Question Regarding ERMS memcpy Linus Torvalds <torvalds@linux-foundation.org> - 2017-03-05 21:20 +0100
                Re: Question Regarding ERMS memcpy Logan Gunthorpe <logang@deltatee.com> - 2017-03-06 08:10 +0100
                Re: Question Regarding ERMS memcpy "H. Peter Anvin" <hpa@zytor.com> - 2017-03-06 08:40 +0100
                Re: Question Regarding ERMS memcpy Logan Gunthorpe <logang@deltatee.com> - 2017-03-06 18:50 +0100
                Re: Question Regarding ERMS memcpy hpa@zytor.com - 2017-03-06 20:10 +0100
                Re: Question Regarding ERMS memcpy Borislav Petkov <bp@suse.de> - 2017-03-06 14:40 +0100
                Re: Question Regarding ERMS memcpy hpa@zytor.com - 2017-03-06 14:50 +0100
                Re: Question Regarding ERMS memcpy Borislav Petkov <bp@suse.de> - 2017-03-06 15:10 +0100
            Re: Question Regarding ERMS memcpy Logan Gunthorpe <logang@deltatee.com> - 2017-03-05 06:00 +0100
              Re: Question Regarding ERMS memcpy Borislav Petkov <bp@suse.de> - 2017-03-05 11:00 +0100

csiph-web