diff options
author | Matthew Auld <matthew.auld@intel.com> | 2022-03-01 11:03:59 +0000 |
---|---|---|
committer | Matthew Auld <matthew.auld@intel.com> | 2022-03-04 15:58:19 +0000 |
commit | 506a890aecb62382a114f43e30b35d96a03fd9ef (patch) | |
tree | 15a69955dd735e53d05a5de56315d31710cfcc8d /tests/i915/gem_exec_params.c | |
parent | 698695136f8ade2391f2d8f45300eae2df02e947 (diff) |
tests/i915/gem_exec_params: check available memory earlier
The shmem mmap and pwrite interfaces conveniently let us probe just a
few pages, without needing to populate the entire object. On discrete
and newer platforms the kernel has dropped support for both, leaving us
with MMAP_OFFSET, which will always populate the entire object, for now
at least. Luckily we can just move the batch creation to after checking
the available memory to ensure we don't hit -ENOMEM on such platforms.
Also it seems that doing a massive allocation(filling much of system
memory) and then calling intel_purge_vm_caches() seems to take 40+
seconds, like when calling intel_require_memory(). Hence switching the
ordering here should also help with that.
For reference the larger-than-life test is just a simple regression test
to ensure that some very large batch buffer(greater than ~4G) can't
overflow the batch_len, causing all kinds of issues. See 57b2d834bf23
("drm/i915/gem: Support parsing of oversize batches").
Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Cc: Ashutosh Dixit <ashutosh.dixit@intel.com>
Cc: Matt Roper <matthew.d.roper@intel.com>
Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com>
Diffstat (limited to 'tests/i915/gem_exec_params.c')
-rw-r--r-- | tests/i915/gem_exec_params.c | 6 |
1 files changed, 3 insertions, 3 deletions
diff --git a/tests/i915/gem_exec_params.c b/tests/i915/gem_exec_params.c index d247b8a6..da36af74 100644 --- a/tests/i915/gem_exec_params.c +++ b/tests/i915/gem_exec_params.c @@ -332,9 +332,7 @@ static void test_larger_than_life_batch(int fd) const struct intel_execution_engine2 *e; uint64_t size = 1ULL << 32; /* batch_len is __u32 as per the ABI */ const intel_ctx_t *ctx = intel_ctx_create_all_physical(fd); - struct drm_i915_gem_exec_object2 exec = { - .handle = batch_create_size(fd, size), - }; + struct drm_i915_gem_exec_object2 exec = {}; struct drm_i915_gem_execbuffer2 execbuf = { .buffers_ptr = to_user_pointer(&exec), .buffer_count = 1, @@ -350,6 +348,8 @@ static void test_larger_than_life_batch(int fd) igt_require(size < gem_aperture_size(fd)); intel_require_memory(2, size, CHECK_RAM); /* batch + shadow */ + exec.handle = batch_create_size(fd, size); + for_each_ctx_engine(fd, ctx, e) { /* Keep the batch_len implicit [0] */ execbuf.flags = e->flags; |