[Intel-gfx] [PATCH i-g-t] tests/gem_buffered_svm: Buffered SVM tests
Belgaumkar, Vinay
vinay.belgaumkar at intel.com
Wed Mar 9 02:00:58 UTC 2016
-----Original Message-----
From: Tvrtko Ursulin [mailto:tvrtko.ursulin at linux.intel.com]
Sent: Wednesday, February 24, 2016 5:18 AM
To: Belgaumkar, Vinay <vinay.belgaumkar at intel.com>; intel-gfx at lists.freedesktop.org
Subject: Re: [Intel-gfx] [PATCH i-g-t] tests/gem_buffered_svm: Buffered SVM tests
Hi,
On 10/02/16 18:01, Vinay Belgaumkar wrote:
> These tests were initially reviewed/merged under the gem_softpin title.
> They use softpinning and userptr mechanism to share buffers between
> CPU and GPU.
>
> The userptr part was decoupled from them recently. Adding these tests
> under a different name to ensure buffered SVM usage testing.
>
> The only change made was to instantiate the drm fd in the main instead
> of every subtest.
>
> Cc: Michel Thierry <michel.thierry at intel.com>
> Cc: Tvrtko Ursulin <tvrtko.ursulin at intel.com>
> ---
> tests/Makefile.sources | 1 +
> tests/gem_buffered_svm.c | 1051 ++++++++++++++++++++++++++++++++++++++++++++++
> 2 files changed, 1052 insertions(+)
> create mode 100644 tests/gem_buffered_svm.c
>
> diff --git a/tests/Makefile.sources b/tests/Makefile.sources index
> df92586..e6ec6f8 100644
> --- a/tests/Makefile.sources
> +++ b/tests/Makefile.sources
> @@ -17,6 +17,7 @@ TESTS_progs_M = \
> drv_hangman \
> gem_bad_reloc \
> gem_basic \
> + gem_buffered_svm \
> gem_busy \
> gem_caching \
> gem_close_race \
> diff --git a/tests/gem_buffered_svm.c b/tests/gem_buffered_svm.c new
> file mode 100644 index 0000000..90e63c4
> --- /dev/null
> +++ b/tests/gem_buffered_svm.c
> @@ -0,0 +1,1051 @@
> +/*
> + * Copyright © 2015 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.
> + *
> + * Authors:
> + * Vinay Belgaumkar <vinay.belgaumkar at intel.com>
> + * Thomas Daniel <thomas.daniel at intel.com>
> + *
> + */
> +
> +#include <unistd.h>
> +#include <stdlib.h>
> +#include <stdint.h>
> +#include <stdio.h>
> +#include <string.h>
> +#include <fcntl.h>
> +#include <inttypes.h>
> +#include <errno.h>
> +#include <sys/stat.h>
> +#include <sys/ioctl.h>
> +#include <sys/time.h>
> +#include <malloc.h>
> +#include "drm.h"
> +#include "ioctl_wrappers.h"
> +#include "drmtest.h"
> +#include "intel_chipset.h"
> +#include "intel_io.h"
> +#include "i915_drm.h"
> +#include <assert.h>
> +#include <sys/wait.h>
> +#include <sys/ipc.h>
> +#include <sys/shm.h>
> +#include "igt_kms.h"
> +#include <inttypes.h>
> +#include <sys/types.h>
> +#include <sys/stat.h>
> +#include "igt.h"
> +
> +#define BO_SIZE 4096
> +#define MULTIPAGE_BO_SIZE 2 * BO_SIZE #define STORE_BATCH_BUFFER_SIZE
> +4
> +#define EXEC_OBJECT_PINNED (1<<4)
> +#define EXEC_OBJECT_SUPPORTS_48B_ADDRESS (1<<3) #define
> +SHARED_BUFFER_SIZE 4096
> +
> +typedef struct drm_i915_gem_userptr i915_gem_userptr;
> +
> +static uint32_t init_userptr(int fd, i915_gem_userptr *, void *ptr,
> +uint64_t size); static void *create_mem_buffer(uint64_t size); static
> +int gem_call_userptr_ioctl(int fd, i915_gem_userptr *userptr); static
> +void gem_pin_userptr_test(int fd); static void gem_pin_bo_test(int
> +fd); static void gem_pin_invalid_vma_test(int fd, bool
> +test_decouple_flags, bool test_canonical_offset); static void
> +gem_pin_overlap_test(int fd); static void
> +gem_pin_high_address_test(int fd);
> +
> +#define NO_PPGTT 0
> +#define ALIASING_PPGTT 1
> +#define FULL_32_BIT_PPGTT 2
> +#define FULL_48_BIT_PPGTT 3
> +/* uses_full_ppgtt
> + * Finds supported PPGTT details.
> + * @fd DRM fd
> + * @min can be
> + * 0 - No PPGTT
> + * 1 - Aliasing PPGTT
> + * 2 - Full PPGTT (32b)
> + * 3 - Full PPGTT (48b)
> + * RETURNS true/false if min support is present */ static bool
> +uses_full_ppgtt(int fd, int min) {
> + struct drm_i915_getparam gp;
> + int val = 0;
> +
> + memset(&gp, 0, sizeof(gp));
> + gp.param = 18; /* HAS_ALIASING_PPGTT */
> + gp.value = &val;
> +
> + if (drmIoctl(fd, DRM_IOCTL_I915_GETPARAM, &gp))
> + return 0;
> +
> + errno = 0;
> + return val >= min;
> +}
gem_gtt_type got added to igt in the meantime so this can be reduced to:
static bool uses_full_ppgtt(int fd, int min) {
return gem_gtt_type(fd) >= min;
}
>> Done.
> +
> +/* gem_call_userptr_ioctl
> + * Helper to call ioctl - TODO: move to lib
> + * @fd - drm fd
> + * @userptr - pointer to initialised userptr
> + * RETURNS status of ioctl call
> +*/
> +static int gem_call_userptr_ioctl(int fd, i915_gem_userptr *userptr)
> +{
> + int ret;
> +
> + ret = drmIoctl(fd, DRM_IOCTL_I915_GEM_USERPTR, userptr);
> +
> + if (ret)
> + ret = errno;
> +
> + return ret;
> +}
> +
> +/* init_userptr
> + * Helper that inits userptr an returns handle
> + * @fd - drm fd
> + * @userptr - pointer to empty userptr
> + * @ptr - buffer to be shared
> + * @size - size of buffer
> + * @ro - read only flag
> + * RETURNS handle to shared buffer
> +*/
> +static uint32_t init_userptr(int fd, i915_gem_userptr *userptr, void *ptr,
> + uint64_t size)
> +{
> + int ret;
> +
> + memset((void*)userptr, 0, sizeof(i915_gem_userptr));
> +
> + userptr->user_ptr = (unsigned long)ptr; /* Need the cast to overcome compiler warning */
> + userptr->user_size = size;
> + userptr->flags = 0; /* use synchronized operation */
> +
> + ret = gem_call_userptr_ioctl(fd, userptr);
> + igt_assert_eq(ret, 0);
> +
> + return userptr->handle;
> +}
Could get rid of the the above two by calling gem_userptr instead.
Unless I am missing something it looks compatible. Probably even better to skip rather than fail on kernels with no synchronized userptr support, which the library helper will do for you automatically.
>> Done
> +
> +/* create_mem_buffer
> + * Creates a 4K aligned CPU buffer
> + * @size - size of buffer
> + * RETURNS pointer to buffer of @size */ static void
> +*create_mem_buffer(uint64_t size) {
> + void *addr;
> + int ret;
> +
> + ret = posix_memalign(&addr, 4096, size);
> + igt_assert(ret == 0);
> +
> + return addr;
> +}
> +
> +/* setup_exec_obj
> + * populate exec object
> + * @exec - exec object
> + * @handle - handle to gem buffer
> + * @flags - any flags
> + * @offset - requested VMA
> +*/
> +static void setup_exec_obj(struct drm_i915_gem_exec_object2 *exec,
> + uint32_t handle, uint32_t flags,
> + uint64_t offset)
> +{
> + memset(exec, 0, sizeof(struct drm_i915_gem_exec_object2));
> + exec->handle = handle;
> + exec->flags = flags;
> + exec->offset = offset;
> +}
> +
> +/* gen8_canonical_addr
> + * Used to convert any address into canonical form, i.e. [63:48] == [47].
> + * Based on kernel's sign_extend64 implementation.
> + * @address - a virtual address
> +*/
> +#define GEN8_HIGH_ADDRESS_BIT 47
> +static uint64_t gen8_canonical_addr(uint64_t address) {
> + __u8 shift = 63 - GEN8_HIGH_ADDRESS_BIT;
> + return (__s64)(address << shift) >> shift; }
> +
> +/* gem_store_data_svm
> + * populate batch buffer with MI_STORE_DWORD_IMM command
> + * @fd: drm file descriptor
> + * @cmd_buf: batch buffer
> + * @vaddr: destination Virtual address
> + * @data: data to be store at destination
> + * @end: whether to end batch buffer or not */ static int
> +gem_store_data_svm(int fd, uint32_t *cmd_buf, uint64_t vaddr,
> + uint32_t data, bool end)
> +{
> + int i = 0;
> +
> + cmd_buf[i++] = MI_STORE_DWORD_IMM;
> + cmd_buf[i++] = vaddr & 0xFFFFFFFC;
> + cmd_buf[i++] = (vaddr >> 32) & 0xFFFF; /* bits 32:47 */
> +
> + cmd_buf[i++] = data;
> + if (end) {
> + cmd_buf[i++] = MI_BATCH_BUFFER_END;
> + cmd_buf[i++] = 0;
> + }
> +
> + return(i * sizeof(uint32_t));
> +}
> +
> +/* gem_store_data
> + * populate batch buffer with MI_STORE_DWORD_IMM command
> + * This one fills up reloc buffer as well
> + * @fd: drm file descriptor
> + * @cmd_buf: batch buffer
> + * @data: data to be store at destination
> + * @reloc - relocation entry
> + * @end: whether to end batch buffer or not */ static int
> +gem_store_data(int fd, uint32_t *cmd_buf,
> + uint32_t handle, uint32_t data,
> + struct drm_i915_gem_relocation_entry *reloc,
> + bool end)
> +{
> + int i = 0;
> +
> + cmd_buf[i++] = MI_STORE_DWORD_IMM;
> + cmd_buf[i++] = 0; /* lower 31 bits of 48 bit address - 0 reloc needed */
> + cmd_buf[i++] = 0; /* upper 15 bits of 48 bit address - 0 reloc needed */
> + reloc->offset = 1 * sizeof(uint32_t);
> + reloc->delta = 0;
> + reloc->target_handle = handle;
> + reloc->read_domains = I915_GEM_DOMAIN_RENDER;
> + reloc->write_domain = I915_GEM_DOMAIN_RENDER;
> + reloc->presumed_offset = 0;
> + cmd_buf[i++] = data;
> + if (end) {
> + cmd_buf[i++] = MI_BATCH_BUFFER_END;
> + cmd_buf[i++] = 0;
> + }
> +
> + return (i * sizeof(uint32_t));
> +}
> +
> +/* setup_execbuffer
> + * helper for buffer execution
> + * @execbuf - pointer to execbuffer
> + * @exec_object - pointer to exec object2 struct
> + * @ring - ring to be used
> + * @buffer_count - how manu buffers to submit
> + * @batch_length - length of batch buffer */ static void
> +setup_execbuffer(struct drm_i915_gem_execbuffer2 *execbuf,
> + struct drm_i915_gem_exec_object2 *exec_object,
> + int ring, int buffer_count, int batch_length) {
> + execbuf->buffers_ptr = (unsigned long)exec_object;
> + execbuf->buffer_count = buffer_count;
> + execbuf->batch_start_offset = 0;
> + execbuf->batch_len = batch_length;
> + execbuf->cliprects_ptr = 0;
> + execbuf->num_cliprects = 0;
> + execbuf->DR1 = 0;
> + execbuf->DR4 = 0;
> + execbuf->flags = ring;
> + i915_execbuffer2_set_context_id(*execbuf, 0);
> + execbuf->rsvd2 = 0;
> +}
> +
> +/* submit_and_sync
> + * Helper function for exec and sync functions
> + * @fd - drm fd
> + * @execbuf - pointer to execbuffer
> + * @batch_buf_handle - batch buffer handle */ static void
> +submit_and_sync(int fd, struct drm_i915_gem_execbuffer2 *execbuf,
> + uint32_t batch_buf_handle)
> +{
> + gem_execbuf(fd, execbuf);
> + gem_sync(fd, batch_buf_handle);
> +}
> +
> +/* gem_userptr_sync
> + * helper for syncing to CPU domain - copy/paste from userblit
> + * @fd - drm fd
> + * @handle - buffer handle to sync
> +*/
> +static void gem_userptr_sync(int fd, uint32_t handle) {
> + gem_set_domain(fd, handle, I915_GEM_DOMAIN_CPU,
> +I915_GEM_DOMAIN_CPU); }
> +
> +
> +/* gem_pin_userptr_test
> + * This test will create a shared buffer, and create a command
> + * for GPU to write data in it
> + * CPU will read and make sure expected value is obtained
> + * Malloc a 4K buffer
> + * Share buffer with with GPU by using userptr ioctl
> + * Create batch buffer to write DATA to first dword of buffer
> + * Use 0x1000 address as destination address in batch buffer
> + * Set EXEC_OBJECT_PINNED flag in exec object
> + * Set 'offset' in exec object to 0x1000
> + * Submit execbuffer
> + * Verify value of first DWORD in shared buffer matches DATA */
> +static void gem_pin_userptr_test(int fd) {
> + i915_gem_userptr userptr;
> + uint32_t *shared_buffer;
> + struct drm_i915_gem_execbuffer2 execbuf;
> + struct drm_i915_gem_exec_object2 exec_object2[2];
> + uint32_t batch_buffer[STORE_BATCH_BUFFER_SIZE + 2];
> + uint32_t batch_buf_handle, shared_buf_handle;
> + int ring, len;
> + const uint32_t data = 0x12345678;
> + uint64_t pinning_offset = 0x1000;
> +
> + igt_require(uses_full_ppgtt(fd, ALIASING_PPGTT));
> + batch_buf_handle = gem_create(fd, BO_SIZE);
> +
> + /* create cpu buffer */
> + shared_buffer = create_mem_buffer(BO_SIZE);
> +
> + /* share with GPU */
> + shared_buf_handle = init_userptr(fd, &userptr, shared_buffer,
> + BO_SIZE);
> +
> + /* create command buffer with write command */
> + len = gem_store_data_svm(fd, batch_buffer, pinning_offset, data, true);
> + gem_write(fd, batch_buf_handle, 0, batch_buffer, len);
> +
> + /* submit command buffer */
> + setup_exec_obj(&exec_object2[0], shared_buf_handle,
> + EXEC_OBJECT_PINNED, pinning_offset);
> + setup_exec_obj(&exec_object2[1], batch_buf_handle, 0, 0);
> +
> + ring = I915_EXEC_RENDER;
> +
> + setup_execbuffer(&execbuf, exec_object2, ring, 2, len);
> + submit_and_sync(fd, &execbuf, batch_buf_handle);
> + gem_userptr_sync(fd, shared_buf_handle);
> +
> + /* Check if driver pinned the buffer as requested */
> + igt_fail_on_f(exec_object2[0].offset != pinning_offset,
> + "\nFailed to pin at requested offset");
> + /* check on CPU to see if value changes */
> + igt_fail_on_f(shared_buffer[0] != data,
> + "\nCPU read does not match GPU write,\
> + expected: 0x%x, got: 0x%x\n",
> + data, shared_buffer[0]);
> +
> + gem_close(fd, batch_buf_handle);
> + gem_close(fd, shared_buf_handle);
> + free(shared_buffer);
> +}
> +
> +/* gem_pin_bo
> + * This test will test softpinning of a gem buffer object
> + * Malloc a 4K buffer
> + * Create batch buffer to write DATA to first dword of buffer
> + * Use 0x1000 address as destination address in batch buffer
> + * Set EXEC_OBJECT_PINNED flag in exec object
> + * Set 'offset' in exec object to 0x1000
> + * Submit execbuffer
> + * Verify value pinned offset matches the request */ static void
> +gem_pin_bo_test(int fd)
This test is not so interesting for buffered SVM I think and if you remove it, you could also get rid of the gem_store_data helper.
>> Done.
> +{
> + struct drm_i915_gem_execbuffer2 execbuf;
> + struct drm_i915_gem_exec_object2 exec_object2[2];
> + uint32_t batch_buffer[STORE_BATCH_BUFFER_SIZE + 2];
> + uint32_t batch_buf_handle, unshared_buf_handle;
> + struct drm_i915_gem_relocation_entry reloc[4];
> + int ring, len;
> + uint32_t value;
> + const uint32_t data = 0x12345678;
> + uint64_t pinning_offset = 0x1000;
> +
> + igt_require(uses_full_ppgtt(fd, ALIASING_PPGTT));
> +
> + batch_buf_handle = gem_create(fd, BO_SIZE);
> +
> + /* create gem buffer */
> + unshared_buf_handle = gem_create(fd, BO_SIZE);
> +
> + /* create command buffer with write command */
> + len = gem_store_data(fd, batch_buffer, unshared_buf_handle, data,
> + reloc, true);
> + gem_write(fd, batch_buf_handle, 0, batch_buffer, len);
> +
> + /* submit command buffer */
> + setup_exec_obj(&exec_object2[0], unshared_buf_handle,
> + EXEC_OBJECT_PINNED, pinning_offset);
> + setup_exec_obj(&exec_object2[1], batch_buf_handle, 0, 0);
> + exec_object2[1].relocation_count = 1;
> + exec_object2[1].relocs_ptr = (unsigned long)reloc;
> +
> + ring = I915_EXEC_RENDER;
> +
> + setup_execbuffer(&execbuf, exec_object2, ring, 2, len);
> + submit_and_sync(fd, &execbuf, batch_buf_handle);
> +
> + /* Check if driver pinned the buffer as requested */
> + igt_fail_on_f(exec_object2[0].offset != pinning_offset,
> + "\nFailed to pin at requested offset");
> + gem_read(fd, unshared_buf_handle, 0, (void*)&value, 4);
> + igt_assert(value == data);
> +
> + gem_close(fd, batch_buf_handle);
> + gem_close(fd, unshared_buf_handle);
> +}
> +
> +
> +/* gem_multiple_process_test
> + * Run basic test simultaneously with multiple processes
> + * This will test pinning same VA separately in each process
> +
> + * fork();
> + * Execute basic test in parent/child processes */ #define
> +MAX_NUM_PROCESSES 10
> +
> +static void gem_multiple_process_test(int fd) {
> + igt_require(uses_full_ppgtt(fd, ALIASING_PPGTT));
> +
> + igt_fork(child, MAX_NUM_PROCESSES) {
> + gem_pin_userptr_test(fd);
> + }
> + igt_waitchildren();
> +
> +}
As an observation, above and below are quite similar in that they do the repinning at the same offset. Forking is just serialized by the struct mutex implicitly, while the below is laid out in the code. But it makes sense to keep both I think in case impossible happens and struct mutex goes away one day.
> +
> +/* gem_repin_test
> + * This test tries to repin a buffer at a previously pinned vma
> + * from a different execbuf.
> + * Malloc a 4K buffer
> + * Share buffer with with GPU by using userptr ioctl
> + * Create batch buffer to write DATA to first dword of buffer
> + * Use 0x1000 address as destination address in batch buffer
> + * Set EXEC_OBJECT_PINNED flag in exec object
> + * Set 'offset' in exec object to 0x1000 VMA
> + * Submit execbuffer
> + * Verify value of first DWORD in shared buffer matches DATA
> +
> + * Create second shared buffer
> + * Follow all steps above
> + * Execpt, for offset, use VMA of first buffer above
> + * Submit execbuffer
> + * Verify value of first DWORD in second shared buffer matches DATA
> +*/
> +static void gem_repin_test(int fd)
> +{
> + i915_gem_userptr userptr;
> + i915_gem_userptr userptr1;
> + uint32_t *shared_buffer;
> + uint32_t *shared_buffer1;
> + struct drm_i915_gem_execbuffer2 execbuf;
> + struct drm_i915_gem_exec_object2 exec_object2[2];
> + uint32_t batch_buffer[STORE_BATCH_BUFFER_SIZE + 2];
> + uint32_t batch_buf_handle, shared_buf_handle, shared_buf_handle1;
> + int ring, len;
> + const uint32_t data = 0x12345678;
> + uint64_t pinning_offset = 0x1000;
> +
> + /* Create gem object */
> + igt_require(uses_full_ppgtt(fd, ALIASING_PPGTT));
> +
> + batch_buf_handle = gem_create(fd, BO_SIZE);
> +
> + /* create cpu buffer, set first elements to 0x0 */
> + shared_buffer = create_mem_buffer(BO_SIZE);
> + shared_buffer1 = create_mem_buffer(BO_SIZE);
> + shared_buffer[0] = 0x0;
> + shared_buffer1[0] = 0x0;
> +
> + /* share with GPU and get handles */
> + shared_buf_handle = init_userptr(fd, &userptr, shared_buffer,
> + BO_SIZE);
> + shared_buf_handle1 = init_userptr(fd, &userptr1, shared_buffer1,
> + BO_SIZE);
> +
> + /* create command buffer with write command */
> + len = gem_store_data_svm(fd, batch_buffer, pinning_offset, data, true);
> + gem_write(fd, batch_buf_handle, 0, batch_buffer, len);
> +
> + /* submit command buffer */
> + setup_exec_obj(&exec_object2[0], shared_buf_handle,
> + EXEC_OBJECT_PINNED, pinning_offset);
> + setup_exec_obj(&exec_object2[1], batch_buf_handle, 0, 0);
> +
> + ring = I915_EXEC_RENDER;
> +
> + setup_execbuffer(&execbuf, exec_object2, ring, 2, len);
> + submit_and_sync(fd, &execbuf, batch_buf_handle);
> + gem_userptr_sync(fd, shared_buf_handle);
> +
> + igt_assert(exec_object2[0].offset == pinning_offset);
> + igt_assert(*shared_buffer == data);
> +
> + /* Second buffer */
> + /* create command buffer with write command */
> + pinning_offset = exec_object2[0].offset;
> + len = gem_store_data_svm(fd, batch_buffer, pinning_offset, data, true);
> + gem_write(fd, batch_buf_handle, 0, batch_buffer, len);
> +
> + /* submit command buffer */
> + /* Pin at shared_buffer, not shared_buffer1 */
> + /* We are requesting address where another buffer was pinned previously */
> + setup_exec_obj(&exec_object2[0], shared_buf_handle1,
> + EXEC_OBJECT_PINNED, pinning_offset);
> + setup_exec_obj(&exec_object2[1], batch_buf_handle, 0, 0);
> +
> + ring = I915_EXEC_RENDER;
> +
> + setup_execbuffer(&execbuf, exec_object2, ring, 2, len);
> + submit_and_sync(fd, &execbuf, batch_buf_handle);
> + gem_userptr_sync(fd, shared_buf_handle1);
> +
> + igt_assert(exec_object2[0].offset == pinning_offset);
> + igt_assert(*shared_buffer1 == data);
> +
> + gem_close(fd, batch_buf_handle);
> + gem_close(fd, shared_buf_handle);
> +
> + free(shared_buffer);
> + free(shared_buffer1);
> +}
> +
> +
> +/* gem_repin_overlap_test
> + * This test will attempt to pin two buffers at the same VMA as part of the same
> + execbuffer object
> +
> + * Malloc a 4K buffer
> + * Share buffer with with GPU by using userptr ioctl
> + * Create second shared buffer
> + * Create batch buffer to write DATA to first dword of each buffer
> + * Use same virtual address as destination addresses in batch buffer
> + * Set EXEC_OBJECT_PINNED flag in both exec objects
> + * Set 'offset' in both exec objects to same VMA
> + * Submit execbuffer
> + * Command should return EINVAL, since we are trying to pin to same VMA
> +*/
> +static void gem_pin_overlap_test(int fd)
Could potentially remove this one since gem_softpin covers it.
>> Done
> +{
> + i915_gem_userptr userptr;
> + i915_gem_userptr userptr1;
> + int ret;
> + uint32_t *shared_buffer;
> + uint32_t *shared_buffer1;
> + struct drm_i915_gem_execbuffer2 execbuf;
> + struct drm_i915_gem_exec_object2 exec_object2[3];
> + uint32_t shared_buf_handle, shared_buf_handle1;
> + int ring, len;
> + uint64_t pinning_offset = 0x1000;
> +
> + igt_require(uses_full_ppgtt(fd, ALIASING_PPGTT));
> +
> + shared_buffer = create_mem_buffer(BO_SIZE);
> + shared_buffer1 = create_mem_buffer(BO_SIZE * 2);
> +
> + /* share with GPU */
> + shared_buf_handle = init_userptr(fd, &userptr, shared_buffer,
> + BO_SIZE);
> + shared_buf_handle1 = init_userptr(fd, &userptr1, shared_buffer1,
> + BO_SIZE * 2);
> +
> + /* submit command buffer */
> + setup_exec_obj(&exec_object2[0], shared_buf_handle,
> + EXEC_OBJECT_PINNED, pinning_offset);
> + setup_exec_obj(&exec_object2[1], shared_buf_handle1,
> + EXEC_OBJECT_PINNED, pinning_offset);
> +
> + ring = I915_EXEC_RENDER;
> +
> + setup_execbuffer(&execbuf, exec_object2, ring, 2, len);
> +
> + ret = drmIoctl(fd, DRM_IOCTL_I915_GEM_EXECBUFFER2, &execbuf);
> +
> + /* expect to fail */
> + igt_assert_neq(ret, 0);
> + igt_assert(errno == EINVAL);
> +
> + free(shared_buffer);
> + free(shared_buffer1);
> +}
> +
> +/* gem_softpin_stress_test
> + * Stress test which creates 10K buffers and shares with GPU
> + * Create 100K uint32 buffers of size 4K each
> + * Share with GPU using userptr ioctl
> + * Create batch buffer to write DATA in first element of each buffer
> + * Pin each buffer to varying addresses starting from 0x800000000000 going below
> + * (requires offsets in canonical form)
> + * Execute Batch Buffer on Blit ring STRESS_NUM_LOOPS times
> + * Validate every buffer has DATA in first element
> + * Rinse and Repeat on Render ring
> +*/
> +#define STRESS_NUM_BUFFERS 100000
> +#define STRESS_NUM_LOOPS 100
> +#define STRESS_STORE_COMMANDS 4 * STRESS_NUM_BUFFERS
> +#define STRESS_START_ADDRESS 0x800000000000
> +static void gem_softpin_stress_test(int fd)
> +{
> + i915_gem_userptr userptr;
> + uint32_t **shared_buffer;
> + uint32_t *shared_handle;
> + struct drm_i915_gem_execbuffer2 execbuf;
> + struct drm_i915_gem_exec_object2 *exec_object2;
> + uint32_t *batch_buffer;
> + uint32_t batch_buf_handle;
> + int ring, len;
> + int buf, loop;
> + uint64_t pinning_offset = STRESS_START_ADDRESS;
> +
> + igt_require(uses_full_ppgtt(fd, FULL_48_BIT_PPGTT));
> +
> + /* Allocate blobs for all data structures */
> + shared_handle = calloc(STRESS_NUM_BUFFERS, sizeof(uint32_t));
> + shared_buffer = calloc(STRESS_NUM_BUFFERS, sizeof(uint32_t *));
> + exec_object2 = calloc(STRESS_NUM_BUFFERS + 1,
> + sizeof(struct drm_i915_gem_exec_object2));
> + /* 4 dwords per buffer + 2 for the end of batchbuffer */
> + batch_buffer = calloc(STRESS_STORE_COMMANDS + 2, sizeof(uint32_t));
> + batch_buf_handle = gem_create(fd, (STRESS_STORE_COMMANDS + 2)*4);
> +
> + /* create command buffer with write commands */
> + len = 0;
> + for(buf = 0; buf < STRESS_NUM_BUFFERS; buf++) {
> + shared_buffer[buf] = create_mem_buffer(BO_SIZE);
> + *shared_buffer[buf] = 0xFFFFFFFF;
> +
> + /* share with GPU */
> + shared_handle[buf] = init_userptr(fd, &userptr,
> + shared_buffer[buf],
> + BO_SIZE);
> +
> + setup_exec_obj(&exec_object2[buf], shared_handle[buf],
> + EXEC_OBJECT_PINNED |
> + EXEC_OBJECT_SUPPORTS_48B_ADDRESS,
> + gen8_canonical_addr(pinning_offset));
> + len += gem_store_data_svm(fd, batch_buffer + (len/4),
> + gen8_canonical_addr(pinning_offset),
> + buf, (buf == STRESS_NUM_BUFFERS-1)? \
> + true:false);
> +
> + /* decremental 4K aligned address */
> + pinning_offset -= ALIGN(BO_SIZE, 4096);
> + }
> +
> + /* setup command buffer */
> + gem_write(fd, batch_buf_handle, 0, batch_buffer, len);
> + setup_exec_obj(&exec_object2[STRESS_NUM_BUFFERS], batch_buf_handle,
> + 0, 0);
> +
> + /* We want to run this on BLT ring if possible */
> + if (HAS_BLT_RING(intel_get_drm_devid(fd))) {
> + ring = I915_EXEC_BLT;
> +
> + setup_execbuffer(&execbuf, exec_object2, ring,
> + STRESS_NUM_BUFFERS + 1, len);
> +
> + for (loop = 0; loop < STRESS_NUM_LOOPS; loop++) {
> + submit_and_sync(fd, &execbuf, batch_buf_handle);
> + /* Set pinning offset back to original value */
> + pinning_offset = STRESS_START_ADDRESS;
> + for(buf = 0; buf < STRESS_NUM_BUFFERS; buf++) {
> + gem_userptr_sync(fd, shared_handle[buf]);
> + igt_assert(exec_object2[buf].offset ==
> + gen8_canonical_addr(pinning_offset));
> + igt_fail_on_f(*shared_buffer[buf] != buf, \
> + "Mismatch in buffer %d, iteration %d: 0x%08X\n", \
> + buf, loop, *shared_buffer[buf]);
> + pinning_offset -= ALIGN(BO_SIZE, 4096);
> + }
> + /* Reset the buffer entries for next iteration */
> + for(buf = 0; buf < STRESS_NUM_BUFFERS; buf++) {
> + *shared_buffer[buf] = 0xFFFFFFFF;
> + }
> + }
> + }
> +
> + /* Now Render Ring */
> + ring = I915_EXEC_RENDER;
> + setup_execbuffer(&execbuf, exec_object2, ring,
> + STRESS_NUM_BUFFERS + 1, len);
> + for (loop = 0; loop < STRESS_NUM_LOOPS; loop++) {
> + submit_and_sync(fd, &execbuf, batch_buf_handle);
> + pinning_offset = STRESS_START_ADDRESS;
> + for(buf = 0; buf < STRESS_NUM_BUFFERS; buf++) {
> + gem_userptr_sync(fd, shared_handle[buf]);
> + igt_assert(exec_object2[buf].offset ==
> + gen8_canonical_addr(pinning_offset));
> + igt_fail_on_f(*shared_buffer[buf] != buf, \
> + "Mismatch in buffer %d, \
> + iteration %d: 0x%08X\n", buf, loop, *shared_buffer[buf]);
> + pinning_offset -= ALIGN(BO_SIZE, 4096);
> + }
> + /* Reset the buffer entries for next iteration */
> + for(buf = 0; buf < STRESS_NUM_BUFFERS; buf++) {
> + *shared_buffer[buf] = 0xFFFFFFFF;
> + }
> + }
> +
> + for(buf = 0; buf < STRESS_NUM_BUFFERS; buf++) {
> + gem_close(fd, shared_handle[buf]);
> + free(shared_buffer[buf]);
> + }
> + gem_close(fd, batch_buf_handle);
> +
> + free(shared_handle);
> + free(shared_buffer);
> + free(exec_object2);
> + free(batch_buffer);
> +}
> +
> +/* gem_write_multipage_buffer
> + * Create a buffer spanning multiple pages, and share with GPU.
> + * Write to every element of the buffer
> + * and verify correct contents.
> +
> + * Create 8K buffer
> + * Share with GPU using userptr ioctl
> + * Create batch buffer to write DATA in all elements of buffer
> + * Execute Batch Buffer
> + * Validate every element has DATA
> +*/
> +
> +#define DWORD_SIZE sizeof(uint32_t)
> +#define BB_SIZE ((MULTIPAGE_BO_SIZE / DWORD_SIZE) * STORE_BATCH_BUFFER_SIZE) + 2
> +#define NUM_DWORDS (MULTIPAGE_BO_SIZE/sizeof(uint32_t))
> +static void gem_write_multipage_buffer_test(int fd)
> +{
> + i915_gem_userptr userptr;
> + uint32_t *shared_buffer;
> + uint32_t shared_handle;
> + struct drm_i915_gem_execbuffer2 execbuf;
> + struct drm_i915_gem_exec_object2 exec_object2[2];
> + uint32_t batch_buffer[BB_SIZE];
> + uint32_t batch_buf_handle;
> + int ring, len, j;
> + uint64_t pinning_offset=0x1000;
> + uint64_t vaddr;
> +
> + igt_require(uses_full_ppgtt(fd, ALIASING_PPGTT));
> +
> + batch_buf_handle = gem_create(fd, sizeof(batch_buffer));
> + shared_buffer = create_mem_buffer(MULTIPAGE_BO_SIZE);
> +
> + len = 0;
> + memset(batch_buffer, 0, sizeof(batch_buffer));
> + memset(shared_buffer, 0, MULTIPAGE_BO_SIZE);
> +
> + /* share with GPU */
> + shared_handle = init_userptr(fd, &userptr, shared_buffer,
> + MULTIPAGE_BO_SIZE);
> + setup_exec_obj(&exec_object2[0], shared_handle,
> + EXEC_OBJECT_PINNED, pinning_offset);
> +
> + /* create command buffer with write commands */
> + vaddr = pinning_offset;
> + for(j=0; j< NUM_DWORDS; j++) {
> + len += gem_store_data_svm(fd, batch_buffer + (len/4), vaddr,
> + j,
> + (j == NUM_DWORDS - 1) ? true:false);
> + vaddr += sizeof(shared_buffer[0]); /* 4 bytes */
> + }
> +
> + gem_write(fd, batch_buf_handle, 0, batch_buffer, len);
> +
> + /* submit command buffer */
> + setup_exec_obj(&exec_object2[1], batch_buf_handle, 0, 0);
> +
> + ring = I915_EXEC_RENDER;
> + setup_execbuffer(&execbuf, exec_object2, ring, 2, len);
> + submit_and_sync(fd, &execbuf, batch_buf_handle);
> + gem_userptr_sync(fd, shared_handle);
> +
> + igt_assert(exec_object2[0].offset == pinning_offset);
> + for(j = 0; j < (MULTIPAGE_BO_SIZE/sizeof(uint32_t)); j++) {
> + igt_fail_on_f(shared_buffer[j] != j,
> + "Mismatch in index %d: 0x%08X\n", j, shared_buffer[j]);
> + }
> +
> + gem_close(fd, batch_buf_handle);
> + gem_close(fd, shared_handle);
> +
> + free(shared_buffer);
> +}
> +
> +/* gem_pin_invalid_vma_test
> + * This test will request to pin a shared buffer to an invalid
> + * VMA > 48-bit address if system supports 48B PPGTT; it also
> + * will test that any attempt of using a 48-bit address requires
> + * the SUPPORTS_48B_ADDRESS flag, and that 48-bit address need to be
> + * in canonical form (bits [63:48] == [47]).
> + * If system supports 32B PPGTT, it will test the equivalent invalid VMA
> + * Create shared buffer of size 4K
> + * Try and Pin object to invalid address
> +*/
> +static void gem_pin_invalid_vma_test(int fd, bool test_decouple_flags,
> + bool test_canonical_offset)
> +{
Can be removed I think, should be covered by gem_softpin.
>> Done.
> + i915_gem_userptr userptr;
> + int ret;
> + uint32_t *shared_buffer;
> + struct drm_i915_gem_execbuffer2 execbuf;
> + struct drm_i915_gem_exec_object2 exec_object2[1];
> + uint32_t shared_buf_handle;
> + int ring;
> + uint64_t invalid_address_for_48b = 0x9000000000000; /* 52 bit address */
> + uint64_t noncanonical_address_for_48b = 0xFF0000000000; /* 48 bit address in noncanonical form */
> + uint64_t invalid_address_for_32b = 0x900000000; /* 36 bit address */
> +
> + igt_require(uses_full_ppgtt(fd, FULL_48_BIT_PPGTT) ||
> + uses_full_ppgtt(fd, FULL_32_BIT_PPGTT));
> +
> + shared_buffer = create_mem_buffer(BO_SIZE);
> + *shared_buffer = 0xFFFFFFFF;
> +
> + /* share with GPU */
> + shared_buf_handle = init_userptr(fd, &userptr, shared_buffer, BO_SIZE);
> +
> + if (uses_full_ppgtt(fd, FULL_48_BIT_PPGTT) && test_canonical_offset) {
> + setup_exec_obj(&exec_object2[0], shared_buf_handle,
> + EXEC_OBJECT_PINNED | EXEC_OBJECT_SUPPORTS_48B_ADDRESS,
> + noncanonical_address_for_48b);
> + } else if (uses_full_ppgtt(fd, FULL_48_BIT_PPGTT) && !test_decouple_flags) {
> + setup_exec_obj(&exec_object2[0], shared_buf_handle,
> + EXEC_OBJECT_PINNED | EXEC_OBJECT_SUPPORTS_48B_ADDRESS,
> + invalid_address_for_48b);
> + } else {
> + /* This also fails in 48b without 48B_ADDRESS support flag */
> + setup_exec_obj(&exec_object2[0], shared_buf_handle,
> + EXEC_OBJECT_PINNED, invalid_address_for_32b);
> + }
> +
> + ring = I915_EXEC_RENDER;
> +
> + setup_execbuffer(&execbuf, exec_object2, ring, 1, 0);
> +
> + /* Expect execbuf to fail */
> + ret = drmIoctl(fd,
> + DRM_IOCTL_I915_GEM_EXECBUFFER2,
> + &execbuf);
> +
> + igt_assert(errno == EINVAL);
> + igt_assert_neq(ret, 0);
> +
> + gem_close(fd, shared_buf_handle);
> + free(shared_buffer);
> +}
> +
> +
> +/* gem_pin_high_address_test
> + * This test will create a shared buffer, and create a command
> + * for GPU to write data in it. It will attempt to pin the buffer at address > 32 bits.
> + * CPU will read and make sure expected value is obtained
> +
> + * Malloc a 4K buffer
> + * Share buffer with with GPU by using userptr ioctl
> + * Create batch buffer to write DATA to first dword of buffer
> + * Use virtual address of buffer as 0x1100000000 (> 32 bit)
> + * Set EXEC_OBJECT_PINNED flag in exec object
> + * Set 'offset' in exec object to shared buffer VMA
> + * Submit execbuffer
> + * Verify value of first DWORD in shared buffer matches DATA
> +*/
> +
> +static void gem_pin_high_address_test(int fd)
> +{
> + i915_gem_userptr userptr;
> + uint32_t *shared_buffer;
> + struct drm_i915_gem_execbuffer2 execbuf;
> + struct drm_i915_gem_exec_object2 exec_object2[2];
> + uint32_t batch_buffer[STORE_BATCH_BUFFER_SIZE + 2];
> + uint32_t batch_buf_handle, shared_buf_handle;
> + int ring, len;
> + const uint32_t data = 0x12345678;
> + uint64_t high_address = 0x1111FFFF000; /* 44 bit address */
> +
> + igt_require(uses_full_ppgtt(fd, FULL_48_BIT_PPGTT));
> +
> + batch_buf_handle = gem_create(fd, BO_SIZE);
> +
> + /* create cpu buffer, set to all 0xF's */
> + shared_buffer = create_mem_buffer(BO_SIZE);
> + *shared_buffer = 0xFFFFFFFF;
> +
> + /* share with GPU */
> + shared_buf_handle = init_userptr(fd, &userptr, shared_buffer, BO_SIZE);
> +
> + /* create command buffer with write command */
> + len = gem_store_data_svm(fd, batch_buffer, high_address, data, true);
> + gem_write(fd, batch_buf_handle, 0, batch_buffer, len);
> +
> + /* submit command buffer */
> + setup_exec_obj(&exec_object2[0], shared_buf_handle,
> + EXEC_OBJECT_PINNED | EXEC_OBJECT_SUPPORTS_48B_ADDRESS, high_address);
> + setup_exec_obj(&exec_object2[1], batch_buf_handle, 0, 0);
> +
> + ring = I915_EXEC_RENDER;
> +
> + setup_execbuffer(&execbuf, exec_object2, ring, 2, len);
> + submit_and_sync(fd, &execbuf, batch_buf_handle);
> + gem_userptr_sync(fd, shared_buf_handle);
> +
> + igt_assert(exec_object2[0].offset == high_address);
> + /* check on CPU to see if value changes */
> + igt_fail_on_f(shared_buffer[0] != data,
> + "\nCPU read does not match GPU write, \
> + expected: 0x%x, got: 0x%x\n", data, shared_buffer[0]);
> +
> + gem_close(fd, batch_buf_handle);
> + gem_close(fd, shared_buf_handle);
> + free(shared_buffer);
> +}
> +
> +/* gem_pin_near_48Bit_test
> + * This test will create a shared buffer,
> + * and create a command for GPU to write data in it. It will attempt
> + * to pin the buffer at address > 47 bits <= 48-bit.
> + * CPU will read and make sure expected value is obtained.
> + * Note that we must submit addresses in canonical form, not only
> + * because the addresss will be validated, but also the returned offset
> + * will be in this format.
> +
> + * Malloc a 4K buffer
> + * Share buffer with with GPU by using userptr ioctl
> + * Create batch buffer to write DATA to first dword of buffer
> + * Use virtual address of buffer as range between 47-bit and 48-bit
> + * Set EXEC_OBJECT_PINNED flag in exec object
> + * Set 'offset' in exec object to shared buffer VMA
> + * Submit execbuffer
> + * Verify value of first DWORD in shared buffer matches DATA
> +*/
> +#define BEGIN_HIGH_ADDRESS 0x7FFFFFFFF000
> +#define END_HIGH_ADDRESS 0xFFFFFFFFC000
> +#define ADDRESS_INCREMENT 0x2000000000
> +static void gem_pin_near_48Bit_test(int fd)
> +{
> + i915_gem_userptr userptr;
> + uint32_t *shared_buffer;
> + struct drm_i915_gem_execbuffer2 execbuf;
> + struct drm_i915_gem_exec_object2 exec_object2[2];
> + uint32_t batch_buffer[BO_SIZE];
> + uint32_t batch_buf_handle, shared_buf_handle;
> + int ring, len;
> + const uint32_t data = 0x12345678;
> + uint64_t high_address, can_high_address;
> +
> + igt_require(uses_full_ppgtt(fd, FULL_48_BIT_PPGTT));
> +
> + batch_buf_handle = gem_create(fd, BO_SIZE);
> +
> + /* create cpu buffer, set to all 0xF's */
> + shared_buffer = create_mem_buffer(BO_SIZE);
> + *shared_buffer = 0xFFFFFFFF;
> +
> + /* share with GPU */
> + shared_buf_handle = init_userptr(fd, &userptr, shared_buffer, BO_SIZE);
> +
> + for (high_address = BEGIN_HIGH_ADDRESS; high_address <= END_HIGH_ADDRESS;
> + high_address+=ADDRESS_INCREMENT) {
> + can_high_address = gen8_canonical_addr(high_address);
> + /* create command buffer with write command */
> + len = gem_store_data_svm(fd, batch_buffer, can_high_address,
> + data, true);
> + gem_write(fd, batch_buf_handle, 0, batch_buffer, len);
> + /* submit command buffer */
> + setup_exec_obj(&exec_object2[0], shared_buf_handle,
> + EXEC_OBJECT_PINNED | EXEC_OBJECT_SUPPORTS_48B_ADDRESS,
> + can_high_address);
> + setup_exec_obj(&exec_object2[1], batch_buf_handle, 0, 0);
> +
> + ring = I915_EXEC_RENDER;
> + setup_execbuffer(&execbuf, exec_object2, ring, 2, len);
> + submit_and_sync(fd, &execbuf, batch_buf_handle);
> + gem_userptr_sync(fd, shared_buf_handle);
> +
> + igt_assert(exec_object2[0].offset == can_high_address);
> + /* check on CPU to see if value changes */
> + igt_fail_on_f(shared_buffer[0] != data,
> + "\nCPU read does not match GPU write, expected: 0x%x, \
> + got: 0x%x\n, 0x%"PRIx64"", data, shared_buffer[0], high_address);
> + }
> +
> + gem_close(fd, batch_buf_handle);
> + gem_close(fd, shared_buf_handle);
> + free(shared_buffer);
> +}
> +
> +
> +igt_main
> +{
> + int fd = -1;
> +
> + igt_skip_on_simulation();
> +
> + igt_fixture {
> + fd = drm_open_driver(DRIVER_INTEL);
> + igt_require(gem_has_softpin(fd));
> + }
> +
> + /* All tests need PPGTT support */
> + igt_subtest("gem_pin_userptr")
> + gem_pin_userptr_test(fd);
> + igt_subtest("gem_pin_bo")
> + gem_pin_bo_test(fd);
> + igt_subtest("gem_multiple_process")
> + gem_multiple_process_test(fd);
> + igt_subtest("gem_repin")
> + gem_repin_test(fd);
> + igt_subtest("gem_pin_overlap")
> + gem_pin_overlap_test(fd);
> + igt_subtest("gem_write_multipage_buffer")
> + gem_write_multipage_buffer_test(fd);
> +
> + /* Following tests need 32/48 Bit PPGTT support */
> + igt_subtest("gem_pin_invalid_vma")
> + gem_pin_invalid_vma_test(fd, false, false);
> +
> + /* Following tests need 48 Bit PPGTT support */
> + igt_subtest("gen_pin_noncanonical_high_address")
> + gem_pin_invalid_vma_test(fd, false, true);
> + igt_subtest("gem_pin_high_address_without_correct_flag")
> + gem_pin_invalid_vma_test(fd, true, false);
> + igt_subtest("gem_softpin_stress")
> + gem_softpin_stress_test(fd);
> + igt_subtest("gem_pin_high_address")
> + gem_pin_high_address_test(fd);
> + igt_subtest("gem_pin_near_48Bit")
> + gem_pin_near_48Bit_test(fd);
> +
> + igt_fixture
> + close(fd);
> +
> + igt_exit();
> +}
>
Looks okayish to me. You could tidy the couple things I've noticed, but
either way:
>> I have addressed all your comments in this next patch.
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin at intel.com>
Another thing which I noticed is that all the tests are "write by the
GPU, read by the CPU". Perhaps as a follow up you should add some which
operate in the opposite fashion, if that is interesting for the buffered
SVM use cases. Should be I think.
>> Agreed. Will work on those next.
Regards,
Tvrtko
More information about the Intel-gfx
mailing list