| Message ID | 20170224105853.5717-1-radoslaw.szwichtenberg@intel.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Fri, Feb 24, 2017 at 11:58:52AM +0100, Radoslaw Szwichtenberg wrote: > There is no point in using custom function if there lib counterpart. > > Cc: Arkadiusz Hiler <arkadiusz.hiler@intel.com> > Cc: Michal Winiarski <michal.winiarski@intel.com> > Signed-off-by: Radoslaw Szwichtenberg <radoslaw.szwichtenberg@intel.com> > --- > tests/pm_rps.c | 79 +++++++++++----------------------------------------------- > 1 file changed, 14 insertions(+), 65 deletions(-) > > diff --git a/tests/pm_rps.c b/tests/pm_rps.c > index 5a54d9d..7aa66a8 100644 > --- a/tests/pm_rps.c > +++ b/tests/pm_rps.c > @@ -560,58 +560,18 @@ static void reset_gpu(void) > close(fd); > } > > -static uint32_t *make_busy(int fd, uint32_t handle) > +static void boost_freq(int fd, int *boost_freqs) > { > - const int gen = intel_gen(intel_get_drm_devid(fd)); > - struct drm_i915_gem_exec_object2 obj; > - struct drm_i915_gem_relocation_entry reloc; > - struct drm_i915_gem_execbuffer2 execbuf; > - uint32_t *batch; > - int i; > - > - memset(&execbuf, 0, sizeof(execbuf)); > - execbuf.buffers_ptr = (uintptr_t)&obj; > - execbuf.buffer_count = 1; > - > - memset(&obj, 0, sizeof(obj)); > - obj.handle = handle; > - > - obj.relocs_ptr = (uintptr_t)&reloc; > - obj.relocation_count = 1; > - memset(&reloc, 0, sizeof(reloc)); > - > - batch = gem_mmap__wc(fd, obj.handle, 0, 4096, PROT_WRITE); > - gem_set_domain(fd, obj.handle, > - I915_GEM_DOMAIN_GTT, I915_GEM_DOMAIN_GTT); > - > - reloc.target_handle = obj.handle; /* recurse */ > - reloc.presumed_offset = 0; > - reloc.offset = sizeof(uint32_t); > - reloc.delta = 0; > - reloc.read_domains = I915_GEM_DOMAIN_COMMAND; > - reloc.write_domain = 0; > - > - i = 0; > - batch[i] = MI_BATCH_BUFFER_START; > - if (gen >= 8) { > - batch[i] |= 1 << 8 | 1; > - batch[++i] = 0; > - batch[++i] = 0; > - } else if (gen >= 6) { > - batch[i] |= 1 << 8; > - batch[++i] = 0; > - } else { > - batch[i] |= 2 << 6; > - batch[++i] = 0; > - if (gen < 4) { > - batch[i] |= 1; > - reloc.delta = 1; > - } > - } > - i++; > + int64_t timeout = 1; > + int ring = -1; > + igt_spin_t *load; > > - gem_execbuf(fd, &execbuf); > - return batch; > + load = igt_spin_batch_new(fd, ring, 0); > + gem_wait(fd, load->handle, &timeout); > + read_freqs(boost_freqs); > + dump(boost_freqs); > + igt_spin_batch_end(load); > + igt_spin_batch_free(fd, load); Ok, couldn't spot a problem with the conversion, but a little whitespace wouldn't go amiss. At least a comment to remind us that it is a "feature" of waiting that grants us a bost. igt_spin_batch_end is redundant as _free will do it for you. -Chris
diff --git a/tests/pm_rps.c b/tests/pm_rps.c index 5a54d9d..7aa66a8 100644 --- a/tests/pm_rps.c +++ b/tests/pm_rps.c @@ -560,58 +560,18 @@ static void reset_gpu(void) close(fd); } -static uint32_t *make_busy(int fd, uint32_t handle) +static void boost_freq(int fd, int *boost_freqs) { - const int gen = intel_gen(intel_get_drm_devid(fd)); - struct drm_i915_gem_exec_object2 obj; - struct drm_i915_gem_relocation_entry reloc; - struct drm_i915_gem_execbuffer2 execbuf; - uint32_t *batch; - int i; - - memset(&execbuf, 0, sizeof(execbuf)); - execbuf.buffers_ptr = (uintptr_t)&obj; - execbuf.buffer_count = 1; - - memset(&obj, 0, sizeof(obj)); - obj.handle = handle; - - obj.relocs_ptr = (uintptr_t)&reloc; - obj.relocation_count = 1; - memset(&reloc, 0, sizeof(reloc)); - - batch = gem_mmap__wc(fd, obj.handle, 0, 4096, PROT_WRITE); - gem_set_domain(fd, obj.handle, - I915_GEM_DOMAIN_GTT, I915_GEM_DOMAIN_GTT); - - reloc.target_handle = obj.handle; /* recurse */ - reloc.presumed_offset = 0; - reloc.offset = sizeof(uint32_t); - reloc.delta = 0; - reloc.read_domains = I915_GEM_DOMAIN_COMMAND; - reloc.write_domain = 0; - - i = 0; - batch[i] = MI_BATCH_BUFFER_START; - if (gen >= 8) { - batch[i] |= 1 << 8 | 1; - batch[++i] = 0; - batch[++i] = 0; - } else if (gen >= 6) { - batch[i] |= 1 << 8; - batch[++i] = 0; - } else { - batch[i] |= 2 << 6; - batch[++i] = 0; - if (gen < 4) { - batch[i] |= 1; - reloc.delta = 1; - } - } - i++; + int64_t timeout = 1; + int ring = -1; + igt_spin_t *load; - gem_execbuf(fd, &execbuf); - return batch; + load = igt_spin_batch_new(fd, ring, 0); + gem_wait(fd, load->handle, &timeout); + read_freqs(boost_freqs); + dump(boost_freqs); + igt_spin_batch_end(load); + igt_spin_batch_free(fd, load); } static void waitboost(bool reset) @@ -619,15 +579,9 @@ static void waitboost(bool reset) int pre_freqs[NUMFREQ]; int boost_freqs[NUMFREQ]; int post_freqs[NUMFREQ]; - uint32_t *batch, handle; - int64_t timeout = 1; int fd = drm_open_driver(DRIVER_INTEL); - /* When we wait upon the GPU, we want to temporarily boost it - * to maximum. - */ - load_helper_run(LOW); igt_debug("Apply low load...\n"); @@ -640,15 +594,10 @@ static void waitboost(bool reset) sleep(1); } - igt_debug("Wait for gpu...\n"); - handle = gem_create(fd, 4096); - batch = make_busy(fd, handle); - gem_wait(fd, handle, &timeout); - read_freqs(boost_freqs); - dump(boost_freqs); - *batch = MI_BATCH_BUFFER_END; - munmap(batch, 4096); - gem_close(fd, handle); + /* When we wait upon the GPU, we want to temporarily boost it + * to maximum. + */ + boost_freq(fd, boost_freqs); igt_debug("Apply low load again...\n"); sleep(1);
There is no point in using custom function if there lib counterpart. Cc: Arkadiusz Hiler <arkadiusz.hiler@intel.com> Cc: Michal Winiarski <michal.winiarski@intel.com> Signed-off-by: Radoslaw Szwichtenberg <radoslaw.szwichtenberg@intel.com> --- tests/pm_rps.c | 79 +++++++++++----------------------------------------------- 1 file changed, 14 insertions(+), 65 deletions(-)