Re: QIV is faster in Linux

2015-12-14 Thread Ryan Freeman
On Mon, Dec 14, 2015 at 02:20:48PM -0500, Alan Corey wrote:
> Stuck them on devio, simpler than that GUI crap
> 
>   http://devio.us/~ab1jx/files/slowness/
> 
> Outputs from time(1) too
> 
> time qiv dsc_2258.jpg
> 
> Copied and pasted, not redirected
> 
> Debian:
> 0.572u 0.316s 0:03.28 26.8% 0+0k 0+0io 0pf+0w
> OpenBSD:
> 1.920u 0.510s 0:05.97 40.7% 0+0k 182+19io 762pf+0w
> 

I can't speak for qiv as I don't have it installed, but feh opened that image
and scrolled around it very quickly.

11:41 ryan@bofh:~$ time feh dsc_2258.jpg  
0m00.77s real 0m00.37s user 0m00.09s system

Compare that with display(1) from the ImageMagick set that took almost 25
seconds (and caused my audio stream to block a couple times).

11:41 ryan@bofh:~$ time display dsc_2258.jpg   
0m23.23s real 0m18.63s user 0m00.61s system

Even the GIMP did very well here:

11:42 ryan@bofh:~$ time gimp dsc_2258.jpg 
While parsing XMP metadata:
Error on line 43 char 1: End of element  not expected in this 
context

Metadata parasite seems to be corrupt
0m06.90s real 0m03.57s user 0m01.03s system

Perhaps QIV is just doing something silly that OpenBSD exposes through
previously mentioned memory sanitizations.

-ryan

> 
> -- 
> Credit is the root of all evil.  - AB1JX



Re: QIV is faster in Linux

2015-12-14 Thread Alan Corey
If I could get logs from the mplayers would that help?  I tried but it
didn't work so I skipped it.  I don't know how to profile on Linux.
The machine has 4 gigs of RAM.

On 12/14/15, Michael McConville  wrote:
> Alan Corey wrote:
>> I'm thinking this is graphics slowness or filesystem slowness.  Both
>> OpenBSD and Debian both have working mplayers, I don't know if that
>> could log something useful about throughput.
>
> Filesystems slowness seems unlikely to me. I'd guess that it's because
> of the rendering acceleration. Maybe we don't support what Linux uses in
> that case, or there's an ifdef somewhere that gives us a sluggish
> fallback.
>
> It could also be because of a pathological memory allocation pattern.
> They have a much bigger performance impact on OpenBSD because of the
> memory sanitization.
>
> The list goes on. You really can't know without profiling.
>
> That said, remember that things like this are almost always a little
> faster on Linux.
>


-- 
Credit is the root of all evil.  - AB1JX



Re: QIV is faster in Linux

2015-12-14 Thread Michael McConville
Alan Corey wrote:
> If I could get logs from the mplayers would that help?  I tried but it
> didn't work so I skipped it.  I don't know how to profile on Linux.
> The machine has 4 gigs of RAM.

No idea. I don't have time to help with this, but someone else might.



Re: QIV is faster in Linux

2015-12-14 Thread Michael McConville
Alan Corey wrote:
> I'm thinking this is graphics slowness or filesystem slowness.  Both
> OpenBSD and Debian both have working mplayers, I don't know if that
> could log something useful about throughput.

Filesystems slowness seems unlikely to me. I'd guess that it's because
of the rendering acceleration. Maybe we don't support what Linux uses in
that case, or there's an ifdef somewhere that gives us a sluggish
fallback.

It could also be because of a pathological memory allocation pattern.
They have a much bigger performance impact on OpenBSD because of the
memory sanitization.

The list goes on. You really can't know without profiling.

That said, remember that things like this are almost always a little
faster on Linux.



Re: QIV is faster in Linux

2015-12-14 Thread Alan Corey
Stuck them on devio, simpler than that GUI crap

  http://devio.us/~ab1jx/files/slowness/

Outputs from time(1) too

time qiv dsc_2258.jpg

Copied and pasted, not redirected

Debian:
0.572u 0.316s 0:03.28 26.8% 0+0k 0+0io 0pf+0w
OpenBSD:
1.920u 0.510s 0:05.97 40.7% 0+0k 182+19io 762pf+0w


-- 
Credit is the root of all evil.  - AB1JX



Re: QIV is faster in Linux

2015-12-14 Thread Alan Corey
> Can you share a photo taken with your camera and the dmesg of your
> computer?

Sure, but where?  The jpegs are 6 - 15 megs each.  I don't know
anything about pastebin-type services yet except that they exist.

I'm thinking this is graphics slowness or filesystem slowness.  Both
OpenBSD and Debian both have working mplayers, I don't know if that
could log something useful about throughput.

-- 
Credit is the root of all evil.  - AB1JX



Re: QIV is faster in Linux

2015-12-14 Thread Juan Francisco Cantero Hurtado
On Mon, Dec 14, 2015 at 02:20:48PM -0500, Alan Corey wrote:
> Stuck them on devio, simpler than that GUI crap
> 
>   http://devio.us/~ab1jx/files/slowness/
> 
> Outputs from time(1) too
> 
> time qiv dsc_2258.jpg
> 
> Copied and pasted, not redirected
> 
> Debian:
> 0.572u 0.316s 0:03.28 26.8% 0+0k 0+0io 0pf+0w
> OpenBSD:
> 1.920u 0.510s 0:05.97 40.7% 0+0k 182+19io 762pf+0w

Debian uses libjpeg-turbo. We use the traditional libjpeg.

Can you convert the image to png and try qiv again?

I would like to know if the problem is related to libjpeg or something
else.

Mount the partitions with "noatime" and "softdep". You can avoid the
slowdown of the filesystem running "cat yourphoto >/dev/null" just
before of to run qiv.

-- 
Juan Francisco Cantero Hurtado http://juanfra.info



Re: QIV is faster in Linux

2015-12-13 Thread Juan Francisco Cantero Hurtado
On Sun, Dec 13, 2015 at 06:49:49PM -0500, Alan Corey wrote:
> Just an observation on my multiboot laptop.  My current Nikon takes 24
> megapixel images, 6000x4000 and I almost dread looking at them in QIV
> under OpenBSD.  It's not so bad in Debian, same hardware.  I don't
> know how to localize or quantify that.  I guess I'd need to build a
> profiled version of QIV, maybe later.  2 profiled versions.

Can you share a photo taken with your camera and the dmesg of your
computer?


-- 
Juan Francisco Cantero Hurtado http://juanfra.info