https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #40 from Tom Stellard ---
(In reply to Linux User from comment #38)
> > Do you have an X server running?
> I am (on one of GPUs). And I can understand X could be slow, etc and
> adjusting intensity on GPU sharing X server is well
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #39 from Michel Dänzer ---
(In reply to Linux User from comment #36)
> P.P.S. and what about better fan/intensity control?
See bug 73338.
(In reply to Linux User from comment #38)
> P.S. also there is another silly issue. If I
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #38 from Linux User ---
> Do you have an X server running?
I am (on one of GPUs). And I can understand X could be slow, etc and adjusting
intensity on GPU sharing X server is well known thing to me.
But I'm talking about GPU
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #37 from Tom Stellard ---
(In reply to Linux User from comment #36)
> Got similar issues when trying scrypt. I have relatively recent bfgminer (~1
> month old git build) and more or less recent graphics stack: 3.17 kernel,
> mesa
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #36 from Linux User ---
Got similar issues when trying scrypt. I have relatively recent bfgminer (~1
month old git build) and more or less recent graphics stack: 3.17 kernel, mesa
10.4 pre-release from Oibaf PPA and LLVM 3.5.1
I
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #35 from Luke-Jr ---
(In reply to Luke-Jr from comment #34)
> [ 855.257659] radeon :01:00.0: ring 3 stalled for more than 1msec
> [ 855.257670] radeon :01:00.0: GPU lockup
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #34 from Luke-Jr ---
(In reply to Luke-Jr from comment #33)
> Looks like clEnqueueNDRangeKernel is re-compiling (or at least optimising?)
> the kernel, and taking a long time to do so.
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #33 from Luke-Jr ---
Looks like clEnqueueNDRangeKernel is re-compiling (or at least optimising?) the
kernel, and taking a long time to do so. Isn't compiling supposed to be
done *once*
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #32 from Christoph Haag ---
Doesn't work for me either with llvm 216281 and mesa git
Oh and maybe it's not known, at least it wasn't so clear to me: I earlier wrote
how my GPU wouldn't get as warm as I expected nor did it display
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #31 from darkbasic ---
kernel 3.17-rc1 + drm-fixes, llvm git, mesa git
--
You are receiving this mail because:
You are the assignee for the bug.
-- next part --
An HTML attachment was scrubbed...
URL:
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #30 from darkbasic ---
I get no errors but a ridiculous hash rate and a 100% error rate:
1 68.0C | 442.2/440.0/ 0.0kh/s | A:0 R:0+0(none) HW:8/100%
--
You are receiving this mail because:
You are the assignee for the bug.
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #29 from darkbasic ---
I will try, but I have to upgrade my stack first because I switched to fglrx.
--
You are receiving this mail because:
You are the assignee for the bug.
-- next part --
An HTML
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #28 from Tom Stellard ---
Does this work with the latest Mesa and LLVM? I've fixed some bugs recently
that may help.
--
You are receiving this mail because:
You are the assignee for the bug.
-- next part --
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #27 from Christoph Haag ---
(In reply to comment #26)
> Have you verified that scrypt actually works? Do you ever get any accepted
> shares? If it does work, you can close this bug and open another for the
> performance issues.
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #26 from Tom Stellard ---
(In reply to comment #25)
> So with 3.16, newest mesa git, llvm trunk from a few days ago etc. I'm
> getting about 12.0 khash/sec out of my HD 7970M with all the stuff one
> should set:
>
>
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #25 from Christoph Haag ---
So with 3.16, newest mesa git, llvm trunk from a few days ago etc. I'm getting
about 12.0 khash/sec out of my HD 7970M with all the stuff one should set:
GPU_MAX_ALLOC_PERCENT=100 GPU_USE_SYNC_OBJECTS=1
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #24 from darkbasic ---
Sorry, the patch previously did nothing because when I upgraded bfgminer I
forgot to add "-S opencl:auto" to the command line parameters.
I applied
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #23 from Christoph Haag ---
Oh, didn't realize the one in the archlinux repository was too old.
Anyway, HW errors and % nonces invalid was climbing until 70+% so something was
wrong anyway.
With bfgminer git I first got
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #22 from Alex Deucher ---
If you are using 3.14, you may have to force your performance level to high or
revert this patch for SI parts:
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #21 from Luke-Jr ---
Old versions report the hashrate wrongly. Try latest git, or at least a
maintained version (3.5.x or 3.10.x)
--
You are receiving this mail because:
You are the assignee
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #20 from Christoph Haag ---
@Luke-Jr
Thanks, I thought that didn't make sense, but then I noticed that I wrote 2&>
instead of 2> so it was my fault after all.
It looks like it's actually running now with all your patches. AWESOME!
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #19 from Tom Stellard ---
(In reply to comment #17)
> Created attachment 94100 [details]
> crash with patches from the pyrit bug
>
> Sorry, about the double entries, I don't know why it does that. Maybe it's
> because it's some
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #18 from Luke-Jr ---
Log output is duplicated to stderr if it is redirected. So, if you redirect
stderr to a file, *only* redirect stderr (not stdout+stderr).
--
You are receiving this mail
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #17 from Christoph Haag ---
Created attachment 94100
--> https://bugs.freedesktop.org/attachment.cgi?id=94100=edit
crash with patches from the pyrit bug
Sorry, about the double entries, I don't know why it does that. Maybe it's
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #12 from Tom Stellard ---
(In reply to comment #11)
> Obviously I did it but I see nothing in stderr except the bfgminer output I
> already posted.
Does bfgminer do anything with the patch, or does it declare the GPU sick? If
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #7 from Christoph Haag ---
Yes, I have the same now. Either a bfgminer update or a llvm/mesa update fixed
the crashes.
It seems like it is trying to use the correct GPU:
[2014-02-14 15:31:32] OCL 0: Attempting to restart
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #6 from darkbasic ---
Forgot to say: card is 7950 (radeonsi)
--
You are receiving this mail because:
You are the assignee for the bug.
-- next part --
An HTML attachment was scrubbed...
URL:
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #5 from darkbasic ---
It doesn't crash for me but it doesn't work at all:
bfgminer version 3.5.7 - Started: [2014-02-14 14:46:50] - [ 0 days 00:01:56]
[M]anage devices [P]ool management [S]ettings [D]isplay options [H]elp [Q]uit
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #4 from Christoph Haag ---
Created attachment 91965
--> https://bugs.freedesktop.org/attachment.cgi?id=91965=edit
R600_DEBUG=cs bfgminer -v1 --intensity 8 --shaders 8192 --scrypt -S opencl:auto
...
This is with upstream llvm
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #3 from Tom Stellard ---
(In reply to comment #0)
> bfgminer --benchmark without scrypt seems to be working fine now.
>
> But I think many people would rather compute scrypt right now on their 7xxx+
> GPUs, just considering the
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #2 from Christoph Haag ---
Created attachment 91500
--> https://bugs.freedesktop.org/attachment.cgi?id=91500=edit
bfgminer with original scrypt kernel and RADEON_DUMP_SHADERS=1 loop error
Well, I would test it if I could get clang
https://bugs.freedesktop.org/show_bug.cgi?id=72785
--- Comment #1 from Christoph Haag ---
Hm, I did think there would be some interest.
So but because the error had something to do with "llvm.SI.loop" and it said it
was in the "search" function there is only this one loop that can be the
32 matches
Mail list logo