| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
| |
Review carefully, it sucks to have to keep track of the number of
command packet dwords emitted in the batch epilogue manually. The
MI_REPORT_PERF_COUNT_BATCH_DWORDS calculation was obviously wrong.
Cc: "12.0 11.1 11.2" <mesa-stable@lists.freedesktop.org>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Clean up misrepetitions ('if if', 'the the' etc) found throughout the
comments. This has been done manually, after grepping
case-insensitively for duplicate if, is, the, then, do, for, an,
plus a few other typos corrected in fly-by
v2:
* proper commit message and non-joke title;
* replace two 'as is' followed by 'is' to 'as-is'.
v3:
* 'a integer' => 'an integer' and similar (originally spotted by
Jason Ekstrand, I fixed a few other similar ones while at it)
Signed-off-by: Giuseppe Bilotta <giuseppe.bilotta@gmail.com>
Reviewed-by: Chad Versace <chad.versace@intel.com>
|
|
|
|
|
|
|
|
|
|
| |
This matches the byte based offset of brw_load_register_mem*.
The function is also moved into intel_batchbuffer.c like
brw_load_register_mem*.
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
|
|
|
|
|
|
|
| |
The next patch will replace the .used field with an on-demand
calculation of batchbuffer usage.
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
|
|
|
|
|
| |
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Both the AMD and Intel APIs provide a dataSize parameter, and this
function would merrily ignore it. Neither API specifies what to do when
the buffer isn't big enough. I take the easy route of writing all the
complete bits of data that will fit. With more complete specs, we could
probably do something different.
I noticed this while looking into an unused parameter warning. The
warning was actually useful!
brw_performance_monitor.c: In function 'brw_get_perf_monitor_result':
brw_performance_monitor.c:1261:37: warning: unused parameter 'data_size' [-Wunused-parameter]
GLsizei data_size,
^
v2: Fix checks to include offset in the calculation. Noticed by Jan.
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Jan Vesely <jan.vesely@rutgers.edu>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
All dd functions take a gl_context as the first parameter. Instead of
removing it, just silence the warning.
brw_performance_monitor.c: In function 'brw_new_perf_monitor':
brw_performance_monitor.c:1354:41: warning: unused parameter 'ctx' [-Wunused-parameter]
brw_new_perf_monitor(struct gl_context *ctx)
^
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Carl Worth <cworth@cworth.org>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
What a useful warning. #ThanksGCC
brw_performance_monitor.c:153:1: warning: 'static' is not at beginning of declaration [-Wold-style-declaration]
const static struct gl_perf_monitor_counter gen5_raw_chaps_counters[] = {
^
brw_performance_monitor.c:185:1: warning: 'static' is not at beginning of declaration [-Wold-style-declaration]
const static int gen5_oa_snapshot_layout[] =
^
brw_performance_monitor.c:221:1: warning: 'static' is not at beginning of declaration [-Wold-style-declaration]
const static struct gl_perf_monitor_group gen5_groups[] = {
^
brw_performance_monitor.c:240:1: warning: 'static' is not at beginning of declaration [-Wold-style-declaration]
const static struct gl_perf_monitor_counter gen6_raw_oa_counters[] = {
^
brw_performance_monitor.c:281:1: warning: 'static' is not at beginning of declaration [-Wold-style-declaration]
const static int gen6_oa_snapshot_layout[] =
^
brw_performance_monitor.c:317:1: warning: 'static' is not at beginning of declaration [-Wold-style-declaration]
const static struct gl_perf_monitor_counter gen6_statistics_counters[] = {
^
brw_performance_monitor.c:332:1: warning: 'static' is not at beginning of declaration [-Wold-style-declaration]
const static int gen6_statistics_register_addresses[] = {
^
brw_performance_monitor.c:346:1: warning: 'static' is not at beginning of declaration [-Wold-style-declaration]
const static struct gl_perf_monitor_group gen6_groups[] = {
^
brw_performance_monitor.c:356:1: warning: 'static' is not at beginning of declaration [-Wold-style-declaration]
const static struct gl_perf_monitor_counter gen7_raw_oa_counters[] = {
^
brw_performance_monitor.c:402:1: warning: 'static' is not at beginning of declaration [-Wold-style-declaration]
const static int gen7_oa_snapshot_layout[] =
^
brw_performance_monitor.c:470:1: warning: 'static' is not at beginning of declaration [-Wold-style-declaration]
const static struct gl_perf_monitor_counter gen7_statistics_counters[] = {
^
brw_performance_monitor.c:493:1: warning: 'static' is not at beginning of declaration [-Wold-style-declaration]
const static int gen7_statistics_register_addresses[] = {
^
brw_performance_monitor.c:515:1: warning: 'static' is not at beginning of declaration [-Wold-style-declaration]
const static struct gl_perf_monitor_group gen7_groups[] = {
^
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Carl Worth <cworth@cworth.org>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
|
|
|
|
| |
Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch reduces the likelihood of pointer arithmetic overflow bugs in
gather_oa_results(), like the one fixed by b69c7c5dac.
I haven't yet encountered any overflow bugs in the wild along this
patch's codepath. But I get nervous when I see code patterns like this:
(void*) + (int) * (int)
I smell 32-bit overflow all over this code.
This patch retypes 'snapshot_size' to 'ptrdiff_t', which should fix any
potential overflow.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Signed-off-by: Chad Versace <chad.versace@linux.intel.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
For a long time, we've wanted a place to put utility code which isn't
directly tied to Mesa or Gallium internals. This patch creates a new
src/util directory for exactly that purpose, and builds the contents as
libmesautil.la.
ralloc seemed like a good first candidate. These days, it's directly
used by mesa/main, i965, i915, and r300g, so keeping it in src/glsl
didn't make much sense.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
v2 (Jason Ekstrand): More realloc uses and some scons fixes
Signed-off-by: Jason Ekstrand <jason.ekstrand@intel.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
|
|
|
|
| |
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
|
|
|
|
|
|
|
|
| |
Check calloc return value and report on error, also later skip
results handling if there was no memory to store results to.
Signed-off-by: Juha-Pekka Heikkila <juhapekka.heikkila@gmail.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
|
|
|
|
| |
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
I guarded half of the callers to start/stop_oa_counters with generation
checks, but missed the other half (which were added later). OACONTROL
doesn't exist on Ironlake, so we better not write it. Also, there's no
need---Ironlake's performance counters are always running.
This patch moves the generation checks into start/stop_oa_counters,
rather than requiring the caller to do them.
Fixes assertion failures in Piglit's AMD_performance_monitor/measure.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Unfortunately, our hardware only has one set of aggregating performance
counters shared between all 3D programs, and their values are not saved
or restored by hardware contexts. Also, at least on Sandybridge and
Ivybridge, the counters lose their values if the GPU goes to sleep.
To work around both of these problems, we have to snapshot the
performance counters at the beginning and end of each batch, similar to
how we handle query objects on platforms that don't support hardware
contexts. I call these "bookend" snapshots.
Since there can be multiple performance monitors active at a time, we
store the bookend snapshots in a global BO, shared by all monitors.
For monitors that span multiple batches, acquiring results involves
adding up three segments:
BeginPerfMonitor --> End of Batch 1 ("head")
Start of Batch 2 --> End of Batch 2
... ("middle")
Start of Batch N-1 --> End of Batch N-1
Start of Batch N --> EndPerfMonitor ("tail")
Monitors that refer to bookend BO snapshots are considered "unresolved".
We delay resolving them (and adding up deltas to obtain the results) as
long as possible to avoid blocking on mapping monitor->oa_bo.
We can also run out of space in the bookend BO, at which point we have
to resolve all unresolved monitors. Then we can throw away the
snapshots and begin writing at the beginning of the buffer.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
|
|
|
|
|
|
|
|
|
| |
Currently, this only considers the monitor start and end snapshots.
This is woefully insufficient, but allows me to add a bunch of the
infrastructure now and flesh it out later.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
We need to start OA at the beginning of each batch where monitors are
active. OACONTROL isn't part of the hardware context, so to avoid
leaving counters enabled for other applications, we turn them off at the
end of the batch too.
We also need to start them at BeginPerfMonitor time (unless they've
already been started). We stop them when the monitor last ends as well.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
|
|
|
|
|
| |
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
|
|
|
|
|
| |
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
MI_REPORT_PERF_COUNT writes a snapshot of the Observability Architecture
counters to a buffer. Exactly how it works varies between generations:
Ironlake requires two packets, Sandybridge has to use GGTT, and Ivybridge
and later use PPGTT.
v2: Assert that we didn't use more space than we reserved (suggested
by Eric Anholt).
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
|
|
|
|
|
|
|
|
|
| |
Using the OA counters requires some per-batch work. When starting and
ending a batch, it's useful to know whether any monitors are actually
interested in OA data.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
In addition to listing the counter names, we include several "remap"
tables. Confusingly, counters are documented with names like "A23",
are written to some buffer offset other than 23, and exposed by core
Mesa under a counter ID that is different still.
The first is inevitable; MI_REPORT_PERF_COUNT writes certain counters to
fixed locations in the buffer. The latter could be avoided, but core
Mesa uses the "Counters" array index as the ID for a counter. We could
do remapping there, but it would just complicate the core Mesa code.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This is fairly simple:
- At BeginPerfMonitor time, take an opening snapshot.
- At EndPerfMonitor time, take a closing snapshot.
- The first time the application asks for results, subtract the two and
store that value. Then free the BO containing the snapshots.
- On subsequent requests for the results, just return the saved value.
- On reset, throw away the results.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
|
|
|
|
|
|
|
|
|
|
|
|
| |
For now, we only support these on Gen6+, since that's what currently
uses hardware contexts. When we add Ironlake hardware context support,
we can add pipeline statistics register support for that as well.
In theory, we could support pipeline statistics counters even without
hardware contexts, but it would be annoyingly painful.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
|
|
|
|
|
|
|
| |
Since we don't support any counters, there are zero groups.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
|
|
|
|
|
|
|
|
|
|
| |
The Observability Architecture counters are 32-bit unsigned values, and
the Pipeline Statistics Register counters are 64-bit unsigned values.
These convenience macros make it easy to create those types of counters.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
|
|
|
|
|
|
|
| |
These stub functions will be filled out in later patches.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This reverts most of commit 0f2da773070c06b6d20ad264d3abb19c4dfd9761.
(I chose to leave the additions to brw_defines.h.)
My previous Ironlake implementation was somewhat broken: counter data
was global, rather than per-context. This meant that performance
monitors captured data from your compositor, 2D driver, and other 3D
programs.
Originally, I believed that Sandybridge and later had an easy way to
avoid this problem (setting per-context flags in OACONTROL), while
Ironlake did not. So I'd intended to leave it as a known limitation of
performance monitoring support on Ironlake. However, this turned out
not to be true.
Unfortunately, our hardware only has one set of aggregating performance
counters shared between all 3D programs, and their values are not saved
or restored by hardware contexts. Also, at least on Sandybridge and
Ivybridge, the counters lose their values if the GPU goes to sleep.
To work around both of these problems, we have to snapshot the
performance counters at the beginning and end of each batch, similar to
how we handle query objects on platforms that don't support hardware
contexts.
For occlusion queries, this batch bookending approach is fairly simple:
only one occlusion query can be active at a time, and the result is a
single integer. Performance monitors are more complex: an arbitrary
number of monitors can be active at a time, each monitoring some subset
of our ~30 observability counters. Individual monitors can be started
and stopped at any point during the batch. Tracking where each monitor
started/ended relative to batch flushes ends up being a pain. And you
can run out of space in the buffer.
Properly supporting this required some serious rearchitecting of the
code. Rather than writing patches to try and morph a broken system into
a working one (which operates quite differently), I decided it would be
simplest to revert the old code and start fresh. Parts will look
familiar, but other parts are new.
I also decided it would be best to include Sandybridge and Ivybridge
support from the start, since the newer platforms have added complexity
that I wanted to make sure worked. They're also what most people care
about these days.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
|
|
Ironlake's counters are always enabled; userspace can simply send a
MI_REPORT_PERF_COUNT packet to take a snapshot of them. This makes it
easy to implement.
The counters are documented in the source code for the intel-gpu-tools
intel_perf_counters utility.
v2: Adjust for core data structure changes. Add a table mapping buffer
object offsets to exposed counters (which changes each generation).
Finally, add report ID assertions to sanity check the BO layout
(thanks to Carl Worth).
v3: Update for core BeginPerfMonitor hook changes (requested by Brian).
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
|