Oops, my mistake for copying a compile fix from another machine...
If we can't detect how much RAM we have, 0 is not a suitable default, so
keep the error for the time being and only downgrade the error for
unknown swap.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
On SandyBridge, the BLT commands were split from the RENDER commands as
well as the BSD split inherited from Ironlake. So we need to make sure
we do exercise each ring, and in order to do so we also need to make
sure each batch takes longer to execute than it takes for us to
submit it.
v2: Exercise each ring sequentially.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
I was interested in finding why my IVB system is not getting GPU turbo
after suspend/resume. The piece that looks weird to me is that
INTERRUPT_THRESHOLD is sitting at 0, whereas pre-suspend it's
0x12000000.
On my system, sys/fcntl.h contains exactly one line:
#include <fcntl.h>
So there's really no need to #ifdef it. Also, intel_mmio.c already
included <fcntl.h>; there's no need to include it twice.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Added the Android.mk file as per Android make system. Also had to modify
the headers with the proper location for fcntl.h as per the Android
bionic headers location
Signed-off-by: Sateesh Kavuri <sateesh.kavuri@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
There may be some updates required, but assuming Ivybridge is similar to
Sandybridge is a decent start; previously it fell through to the Gen2/3
case and nothing worked.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Ben Widawsky <ben@bwidawsk.net>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
In this case, the code already has a comment about the problem. We don't
need to flood the build with this less than useful message.
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Some test that trash the aperture necessarily need to take forever, so
add a little progress indicator to keep worried minds at peace.
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Lots of tests need to create havoc to LRUs in the kernel or otherwise
need to shuffle things around a bit. So make a small array permutation
function available.
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Test that try to exercise the swap paths need to check whether swap is
available, otherwise they'll just oom.
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Headers are found under top_srcdir/...
Headers are CPP flags, not C Flags
Acked-by: Ben Widawsky <ben@bwidawsk.net>
Signed-off-by: Gaetan Nadon <memsize@videotron.ca>
Some tests are higly timing dependent and others carelessly leave
active buffers behind. So add a helper to quiescent the gpu and call
it unconditionally when opening an fd in a vain attempt to make all
this race-condition hitting more scientifically sound.
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Way too much copy-pasting going on here.
Also fix a compiler warnings in gem_stress while fixup things up.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
<fcntl.h> is the POSIX standard specified header for open(), so should
be available on all platforms.
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Use CWARNFLAGS as in all of xorg. There seems to be no reason why this
module should be different. The warnings were updated recently
for those who install the latest util-macros.
Signed-off-by: Gaetan Nadon <memsize@videotron.ca>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Right now, we only check for hardware DRRS support. But much more can be
done with it. Some day.
Signed-off-by: Eugeni Dodonov <eugeni.dodonov@intel.com>
