On Wed, Aug 03, 2022 at 05:28:16PM +0200, Michael Niedermayer wrote:
> On Tue, Aug 02, 2022 at 11:30:39AM +0200, Anton Khirnov wrote:
> > Quoting Michael Niedermayer (2022-07-30 19:12:39)
> > > On Sat, Jul 30, 2022 at 11:55:45AM +0200, Andreas Rheinhardt wrote:
> > > > Michael Niedermayer:
> > > > > On Wed, Jul 27, 2022 at 12:08:09AM +0200, Andreas Rheinhardt wrote:
> > > > >> Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
> > > > >> ---
> > > > >>  libavcodec/arm/mpegvideoencdsp_init_arm.c  |  4 ++--
> > > > >>  libavcodec/mips/h263dsp_mips.h             |  2 +-
> > > > >>  libavcodec/mips/mpegvideoencdsp_msa.c      |  4 ++--
> > > > >>  libavcodec/mpegvideoencdsp.c               | 10 +++++-----
> > > > >>  libavcodec/mpegvideoencdsp.h               | 10 +++++-----
> > > > >>  libavcodec/ppc/mpegvideoencdsp.c           |  8 ++++----
> > > > >>  libavcodec/x86/mpegvideoenc_qns_template.c |  4 ++--
> > > > >>  libavcodec/x86/mpegvideoencdsp.asm         |  4 ++--
> > > > >>  libavcodec/x86/mpegvideoencdsp_init.c      |  6 +++---
> > > > >>  9 files changed, 26 insertions(+), 26 deletions(-)
> > > > > 
> > > > > LGTM, also all similar things should be ok
> > > > > 
> > > > > [...]
> > > > > 
> > > > 
> > > > Unfortunately, it is not so easy. The first version of this patchset
> > > > (this here) forgot to update the definitions of several loongarch
> > > > functions, leading to lots of red on patchwork. The second version fixed
> > > > the compilation error, but forgot some constifications due to
> > > > unfortunate oversights. For the third version, I wanted to actually
> > > > compile loongarch instead of just staring at the code; but the
> > > > GCC-cross-compiler does not seem to support lsx and lasx (and looking at
> > > > the loongarch-code in GCC makes me believe that they are just not
> > > > supported (yet)). Therefore I resorted to creating dummy-functions for
> > > > all the __lasx_* and __lsx_* functions (they show no more
> > > > const-warnings). I wanted to let the FATE-loongarch-box test this once
> > > > more (and tell me whether there are still any issues left), yet this box
> > > > is no longer active since a few days (for reasons unknown). Can you
> > > > compile the lsx and lasx code (after all, you committed all the
> > > > loongarch code)?
> > > 
> > > i have no easy way to test loongarch, what i can test is qemu-mips.
> > > i think i have said that a few times already.
> > > 
> > > I do have a loongson box that is IIRC ready to be shipped (within EU) to
> > > whoever volunteers to maintain that box for FFmpeg. I dont know if that
> > > box would help with the current code. The OS will liklely need to be updated
> > > and IIRC this has no normal update mechanism) I didnt touch this box for a
> > > really long time.
> > 
> > What OS is it?
> 
> looking thorugh past emails (so i dont have to get the box and unpack it again)
> 
> its some custom fedora/redhat
> that additional info i found while double checking the OS:
> 
>  according to /proc/cpuinfo, 4 core
>  system type             : lst-loongson3-eva
>  machine                 : Unknown
>  processor               : 0
>  cpu model               : ICT Loongson-3A V0.5  FPU V0.1
>  BogoMIPS                : 717.28
>  ...
>  with 500gb 7200rpm WD Caviar Blue HDD
>  4gb memory
> 
> 
> > 
> > I could conceivably host such a machine and give other devs SSH access
> > to it, if it does not require too much maintenance and/or electricity.
> 
> given the tiny heatsink/fan on the CPU i dont think it would require alot of power
> send me teh target address and ill get it shipped to you assuming the box is still where
> it was and i find it

more precisse power usage: (from 2015)

~1W when OFF
~44W when IDLE
~48W when running fate tests

in 2015 with FFmpeg from 2015:

4m49.192s for time ./configure --enable-gpl --enable-pthreads --samples=/home/loongson/fate/ --enable-nonfree --enable-version3 --assert-level=2 --cpu=loongson3a --enable-loongson3
19m31.114s for make -j4

time make fate -j4 -k >& fatelist8
real 28m32.631s
user 98m50.422s
sys 12m53.547s

CPU coolong fan was also replaced with a Noctua NF-A6x25 FLX as the original was really loud

thx

[...]
-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

Elect your leaders based on what they did after the last election, not
based on what they say before an election.