Quoting Mika Kuoppala (2019-11-13 15:49:13)
> From: Imre Deak <imre.deak@xxxxxxxxx>
>
> Add a test to exercise the kernel's mechanism to detection of RC6
> context corruptions, take the necessary action in response (disable
> RC6 and runtime PM) and recover when possible (after system
> suspend/resume).
>
> v2:
> - Skip test on non-existing engines.
> - Fix for old kernels where the command parser returned EINVAL
> instead of EACCESS for a banned privilidged command.
>
> Signed-off-by: Imre Deak <imre.deak@xxxxxxxxx>
> ---
> tests/Makefile.sources | 3 +
> tests/i915/i915_rc6_ctx_corruption.c | 196 +++++++++++++++++++++++++++
> tests/meson.build | 1 +
> 3 files changed, 200 insertions(+)
> create mode 100644 tests/i915/i915_rc6_ctx_corruption.c
>
> diff --git a/tests/Makefile.sources b/tests/Makefile.sources
> index 75c79edb..c0f401c7 100644
> --- a/tests/Makefile.sources
> +++ b/tests/Makefile.sources
> @@ -529,6 +529,9 @@ i915_pm_sseu_SOURCES = i915/i915_pm_sseu.c
> TESTS_progs += i915_query
> i915_query_SOURCES = i915/i915_query.c
>
> +TESTS_progs += i915_rc6_ctx_corruption
> +i915_rc6_ctx_corruption_SOURCES = i915/i915_rc6_ctx_corruption.c
> +
> TESTS_progs += i915_selftest
> i915_selftest_SOURCES = i915/i915_selftest.c
>
> diff --git a/tests/i915/i915_rc6_ctx_corruption.c b/tests/i915/i915_rc6_ctx_corruption.c
> new file mode 100644
> index 00000000..a3326307
> --- /dev/null
> +++ b/tests/i915/i915_rc6_ctx_corruption.c
> @@ -0,0 +1,196 @@
> +/*
> + * Copyright © 2019 Intel Corporation
> + *
> + * Permission is hereby granted, free of charge, to any person obtaining a
> + * copy of this software and associated documentation files (the "Software"),
> + * to deal in the Software without restriction, including without limitation
> + * the rights to use, copy, modify, merge, publish, distribute, sublicense,
> + * and/or sell copies of the Software, and to permit persons to whom the
> + * Software is furnished to do so, subject to the following conditions:
> + *
> + * The above copyright notice and this permission notice (including the next
> + * paragraph) shall be included in all copies or substantial portions of the
> + * Software.
> + *
> + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
> + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
> + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
> + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
> + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
> + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
> + * IN THE SOFTWARE.
> + */
> +
> +#include <stdint.h>
> +
> +#include "igt.h"
> +#include "igt_gt.h"
> +#include "igt_device.h"
> +
> +#define CTX_INFO_REG 0xA300
> +
> +static int drm_fd;
> +static struct pci_device *pci_dev;
> +static uint32_t devid;
> +
> +static int read_reg_on_engine(const struct intel_execution_engine2 *engine,
> + int reg_address, uint32_t *reg_val)
> +{
> + const bool r64b = intel_gen(devid) >= 8;
> + struct drm_i915_gem_exec_object2 obj[2];
> + struct drm_i915_gem_relocation_entry reloc[1];
> + struct drm_i915_gem_execbuffer2 execbuf;
> + uint32_t *batch;
> + uint32_t *dst_buf;
> + int ret;
> + int i;
> +
> + memset(obj, 0, sizeof(obj));
> + obj[0].handle = gem_create(drm_fd, 4096);
> + obj[1].handle = gem_create(drm_fd, 4096);
> + obj[1].relocs_ptr = to_user_pointer(reloc);
> + obj[1].relocation_count = 1;
> +
> + batch = gem_mmap__cpu(drm_fd, obj[1].handle, 0, 4096, PROT_WRITE);
> + gem_set_domain(drm_fd, obj[1].handle,
> + I915_GEM_DOMAIN_CPU, I915_GEM_DOMAIN_CPU);
> +
> + i = 0;
> + batch[i++] = 0x24 << 23 | (1 + r64b); /* SRM */
> + batch[i++] = reg_address;
> + reloc[0].target_handle = obj[0].handle;
> + reloc[0].presumed_offset = obj[0].offset;
> + reloc[0].offset = i * sizeof(uint32_t);
> + reloc[0].delta = 0;
> + reloc[0].read_domains = I915_GEM_DOMAIN_RENDER;
> + reloc[0].write_domain = I915_GEM_DOMAIN_RENDER;
> + batch[i++] = reloc[0].delta;
> + if (r64b)
> + batch[i++] = 0;
> +
> + batch[i++] = MI_BATCH_BUFFER_END;
> + munmap(batch, 4096);
> +
> + memset(&execbuf, 0, sizeof(execbuf));
> + execbuf.buffers_ptr = to_user_pointer(obj);
> + execbuf.buffer_count = 2;
> + execbuf.flags = engine->flags;
> +
> + ret = __gem_execbuf(drm_fd, &execbuf);
> + gem_close(drm_fd, obj[1].handle);
> +
> + if (ret)
> + goto err;
> +
> + if (reg_val) {
> + dst_buf = gem_mmap__cpu(drm_fd, obj[0].handle, 0, 4096,
> + PROT_READ);
> + gem_set_domain(drm_fd, obj[0].handle, I915_GEM_DOMAIN_CPU, 0);
> +
> + *reg_val = dst_buf[0];
> + munmap(dst_buf, 4096);
> + }
> +
> +err:
> + gem_close(drm_fd, obj[0].handle);
> +
> + return ret;
> +}
> +
> +static bool need_ctx_corruption_wa(void)
> +{
> + return IS_BROADWELL(devid) || intel_gen(devid) == 9;
That's a little self servicing and doesn't help find bugs where you
weren't expecting. The test would seem to be useless at predicting new
failures.
This should just be a selftest imo.
-Chris
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
