From: Daniel Vetter <daniel@ffwll.ch>
To: Tiago Vignatti <tiago.vignatti@intel.com>
Cc: daniel.thompson@linaro.org, daniel.vetter@ffwll.ch,
intel-gfx@lists.freedesktop.org,
Rob Bradford <rob@linux.intel.com>,
dri-devel@lists.freedesktop.org, sumit.semwal@linaro.org
Subject: Re: [PATCH 1/7] prime_mmap: Add new test for calling mmap() on dma-buf fds
Date: Thu, 13 Aug 2015 09:04:11 +0200 [thread overview]
Message-ID: <20150813070411.GV17734@phenom.ffwll.local> (raw)
In-Reply-To: <1439422160-20148-6-git-send-email-tiago.vignatti@intel.com>
On Wed, Aug 12, 2015 at 08:29:14PM -0300, Tiago Vignatti wrote:
> From: Rob Bradford <rob@linux.intel.com>
>
> This test has the following subtests:
> - test_correct for correctness of the data
> - test_map_unmap checks for mapping idempotency
> - test_reprime checks for dma-buf creation idempotency
> - test_forked checks for multiprocess access
> - test_refcounting checks for buffer reference counting
> - test_dup chats that dup()ing the fd works
> - test_errors checks the error return values for failures
> - test_aperture_limit tests multiple buffer creation at the gtt aperture
> limit
>
> Signed-off-by: Rob Bradford <rob@linux.intel.com>
> Signed-off-by: Tiago Vignatti <tiago.vignatti@intel.com>
> ---
> tests/Makefile.sources | 1 +
> tests/prime_mmap.c | 383 +++++++++++++++++++++++++++++++++++++++++++++++++
> 2 files changed, 384 insertions(+)
> create mode 100644 tests/prime_mmap.c
>
> diff --git a/tests/Makefile.sources b/tests/Makefile.sources
> index c94714b..5b2072e 100644
> --- a/tests/Makefile.sources
> +++ b/tests/Makefile.sources
> @@ -90,6 +90,7 @@ TESTS_progs_M = \
> pm_rps \
> pm_rc6_residency \
> pm_sseu \
> + prime_mmap \
> prime_self_import \
> template \
> $(NULL)
> diff --git a/tests/prime_mmap.c b/tests/prime_mmap.c
> new file mode 100644
> index 0000000..4dc2055
> --- /dev/null
> +++ b/tests/prime_mmap.c
> @@ -0,0 +1,383 @@
> +/*
> + * Copyright © 2014 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:
> + * Rob Bradford <rob at linux.intel.com>
> + *
> + */
> +
> +/*
> + * Testcase: Check whether mmap()ing dma-buf works
> + */
> +#define _GNU_SOURCE
> +#include <unistd.h>
> +#include <stdlib.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 <pthread.h>
> +
> +#include "drm.h"
> +#include "i915_drm.h"
> +#include "drmtest.h"
> +#include "igt_debugfs.h"
> +#include "ioctl_wrappers.h"
> +
> +#define BO_SIZE (16*1024)
> +
> +static int fd;
> +
> +char pattern[] = {0xff, 0x00, 0x00, 0x00,
> + 0x00, 0xff, 0x00, 0x00,
> + 0x00, 0x00, 0xff, 0x00,
> + 0x00, 0x00, 0x00, 0xff};
> +
> +static void
> +fill_bo(uint32_t handle, size_t size)
> +{
> + off_t i;
> + for (i = 0; i < size; i+=sizeof(pattern))
> + {
> + gem_write(fd, handle, i, pattern, sizeof(pattern));
> + }
> +}
> +
> +static int
> +pattern_check(char *ptr, size_t size)
> +{
> + off_t i;
> + for (i = 0; i < size; i+=sizeof(pattern))
> + {
> + if (memcmp(ptr, pattern, sizeof(pattern)) != 0)
> + return 1;
> + }
> +
> + return 0;
> +}
> +
> +static void
> +test_correct(void)
> +{
> + int dma_buf_fd;
> + char *ptr1, *ptr2;
> + uint32_t handle;
> +
> + handle = gem_create(fd, BO_SIZE);
> + fill_bo(handle, BO_SIZE);
> +
> + dma_buf_fd = prime_handle_to_fd(fd, handle);
> + igt_assert(errno == 0);
> +
> + /* Check correctness vs GEM_MMAP_GTT */
> + ptr1 = gem_mmap(fd, handle, BO_SIZE, PROT_READ | PROT_WRITE);
> + ptr2 = mmap(NULL, BO_SIZE, PROT_READ, MAP_SHARED, dma_buf_fd, 0);
> + igt_assert(ptr1 != MAP_FAILED);
> + igt_assert(ptr2 != MAP_FAILED);
> + igt_assert(memcmp(ptr1, ptr2, BO_SIZE) == 0);
> +
> + /* Check pattern correctness */
> + igt_assert(pattern_check(ptr2, BO_SIZE) == 0);
> +
> + munmap(ptr1, BO_SIZE);
> + munmap(ptr2, BO_SIZE);
> + close(dma_buf_fd);
> + gem_close(fd, handle);
> +}
> +
> +static void
> +test_map_unmap(void)
> +{
> + int dma_buf_fd;
> + char *ptr;
> + uint32_t handle;
> +
> + handle = gem_create(fd, BO_SIZE);
> + fill_bo(handle, BO_SIZE);
> +
> + dma_buf_fd = prime_handle_to_fd(fd, handle);
> + igt_assert(errno == 0);
> +
> + ptr = mmap(NULL, BO_SIZE, PROT_READ, MAP_SHARED, dma_buf_fd, 0);
> + igt_assert(ptr != MAP_FAILED);
> + igt_assert(pattern_check(ptr, BO_SIZE) == 0);
> +
> + /* Unmap and remap */
> + munmap(ptr, BO_SIZE);
> + ptr = mmap(NULL, BO_SIZE, PROT_READ, MAP_SHARED, dma_buf_fd, 0);
> + igt_assert(ptr != MAP_FAILED);
> + igt_assert(pattern_check(ptr, BO_SIZE) == 0);
> +
> + munmap(ptr, BO_SIZE);
> + close(dma_buf_fd);
> + gem_close(fd, handle);
> +}
> +
> +/* prime and then unprime and then prime again the same handle */
> +static void
> +test_reprime(void)
> +{
> + int dma_buf_fd;
> + char *ptr;
> + uint32_t handle;
> +
> + handle = gem_create(fd, BO_SIZE);
> + fill_bo(handle, BO_SIZE);
> +
> + dma_buf_fd = prime_handle_to_fd(fd, handle);
> + igt_assert(errno == 0);
> +
> + ptr = mmap(NULL, BO_SIZE, PROT_READ, MAP_SHARED, dma_buf_fd, 0);
> + igt_assert(ptr != MAP_FAILED);
> + igt_assert(pattern_check(ptr, BO_SIZE) == 0);
> +
> + close (dma_buf_fd);
> + igt_assert(pattern_check(ptr, BO_SIZE) == 0);
> + munmap(ptr, BO_SIZE);
> +
> + dma_buf_fd = prime_handle_to_fd(fd, handle);
> + ptr = mmap(NULL, BO_SIZE, PROT_READ, MAP_SHARED, dma_buf_fd, 0);
> + igt_assert(ptr != MAP_FAILED);
> + igt_assert(pattern_check(ptr, BO_SIZE) == 0);
> +
> + munmap(ptr, BO_SIZE);
> + close(dma_buf_fd);
> + gem_close(fd, handle);
> +}
> +
> +/* map from another process */
> +static void
> +test_forked(void)
> +{
> + int dma_buf_fd;
> + char *ptr;
> + uint32_t handle;
> +
> + handle = gem_create(fd, BO_SIZE);
> + fill_bo(handle, BO_SIZE);
> +
> + dma_buf_fd = prime_handle_to_fd(fd, handle);
> + igt_assert(errno == 0);
> +
> + igt_fork(childno, 1) {
> + ptr = mmap(NULL, BO_SIZE, PROT_READ, MAP_SHARED, dma_buf_fd, 0);
> + igt_assert(ptr != MAP_FAILED);
> + igt_assert(pattern_check(ptr, BO_SIZE) == 0);
> + munmap(ptr, BO_SIZE);
> + close(dma_buf_fd);
> + }
> + close(dma_buf_fd);
> + igt_waitchildren();
> + gem_close(fd, handle);
> +}
> +
> +static void
> +test_refcounting(void)
> +{
> + int dma_buf_fd;
> + char *ptr;
> + uint32_t handle;
> +
> + handle = gem_create(fd, BO_SIZE);
> + fill_bo(handle, BO_SIZE);
> +
> + dma_buf_fd = prime_handle_to_fd(fd, handle);
> + igt_assert(errno == 0);
> + /* Close gem object before mapping */
> + gem_close(fd, handle);
> +
> + ptr = mmap(NULL, BO_SIZE, PROT_READ, MAP_SHARED, dma_buf_fd, 0);
> + igt_assert(ptr != MAP_FAILED);
> + igt_assert(pattern_check(ptr, BO_SIZE) == 0);
> + munmap(ptr, BO_SIZE);
> + close (dma_buf_fd);
> +}
> +
> +/* dup before mmap */
> +static void
> +test_dup(void)
> +{
> + int dma_buf_fd;
> + char *ptr;
> + uint32_t handle;
> +
> + handle = gem_create(fd, BO_SIZE);
> + fill_bo(handle, BO_SIZE);
> +
> + dma_buf_fd = dup(prime_handle_to_fd(fd, handle));
> + igt_assert(errno == 0);
> +
> + ptr = mmap(NULL, BO_SIZE, PROT_READ, MAP_SHARED, dma_buf_fd, 0);
> + igt_assert(ptr != MAP_FAILED);
> + igt_assert(pattern_check(ptr, BO_SIZE) == 0);
> + munmap(ptr, BO_SIZE);
> + gem_close(fd, handle);
> + close (dma_buf_fd);
> +}
> +
> +
> +/* Used for error case testing to avoid wrapper */
> +static int prime_handle_to_fd_no_assert(uint32_t handle, int *fd_out)
> +{
> + struct drm_prime_handle args;
> + int ret;
> +
> + args.handle = handle;
> + args.flags = DRM_CLOEXEC;
> + args.fd = -1;
> +
> + ret = drmIoctl(fd, DRM_IOCTL_PRIME_HANDLE_TO_FD, &args);
> +
> + *fd_out = args.fd;
> +
> + return ret;
> +}
> +
> +static void
> +test_errors(void)
> +{
> + int dma_buf_fd;
> + char *ptr;
> + uint32_t handle;
> +
> + /* Close gem object before priming */
> + handle = gem_create(fd, BO_SIZE);
> + fill_bo(handle, BO_SIZE);
> + gem_close(fd, handle);
> + prime_handle_to_fd_no_assert(handle, &dma_buf_fd);
> + igt_assert(dma_buf_fd == -1 && errno == ENOENT);
> + errno = 0;
> +
> + /* close fd before mapping */
> + handle = gem_create(fd, BO_SIZE);
> + fill_bo(handle, BO_SIZE);
> + dma_buf_fd = prime_handle_to_fd(fd, handle);
> + igt_assert(errno == 0);
> + close(dma_buf_fd);
> + ptr = mmap(NULL, BO_SIZE, PROT_READ, MAP_SHARED, dma_buf_fd, 0);
> + igt_assert(ptr == MAP_FAILED && errno == EBADF);
> + errno = 0;
> + gem_close(fd, handle);
> +
> + /* Map too big */
> + handle = gem_create(fd, BO_SIZE);
> + fill_bo(handle, BO_SIZE);
> + dma_buf_fd = prime_handle_to_fd(fd, handle);
> + igt_assert(errno == 0);
> + ptr = mmap(NULL, BO_SIZE * 2, PROT_READ, MAP_SHARED, dma_buf_fd, 0);
> + igt_assert(ptr == MAP_FAILED && errno == EINVAL);
> + errno = 0;
> + close(dma_buf_fd);
> + gem_close(fd, handle);
That only checks for one of the conditions, trying to map something
offset/overlapping the end of the buffer, but small enough needs to be
tested separately.
Also I think a testcase for invalid buffer2fd flags would be good, just
for completeness - we seem to be missing that one.
-Daniel
> +}
> +
> +static void
> +test_aperture_limit(void)
> +{
> + int dma_buf_fd1, dma_buf_fd2;
> + char *ptr1, *ptr2;
> + uint32_t handle1, handle2;
> + /* Two buffers the sum of which > mappable aperture */
> + uint64_t size1 = (gem_mappable_aperture_size() * 7) / 8;
> + uint64_t size2 = (gem_mappable_aperture_size() * 3) / 8;
> +
> + handle1 = gem_create(fd, size1);
> + fill_bo(handle1, BO_SIZE);
> +
> + dma_buf_fd1 = prime_handle_to_fd(fd, handle1);
> + igt_assert(errno == 0);
> + ptr1 = mmap(NULL, size1, PROT_READ, MAP_SHARED, dma_buf_fd1, 0);
> + igt_assert(ptr1 != MAP_FAILED);
> + igt_assert(pattern_check(ptr1, BO_SIZE) == 0);
> +
> + handle2 = gem_create(fd, size1);
> + fill_bo(handle2, BO_SIZE);
> + dma_buf_fd2 = prime_handle_to_fd(fd, handle2);
> + igt_assert(errno == 0);
> + ptr2 = mmap(NULL, size2, PROT_READ, MAP_SHARED, dma_buf_fd2, 0);
> + igt_assert(ptr2 != MAP_FAILED);
> + igt_assert(pattern_check(ptr2, BO_SIZE) == 0);
> +
> + igt_assert(memcmp(ptr1, ptr2, BO_SIZE) == 0);
> +
> + munmap(ptr1, size1);
> + munmap(ptr2, size2);
> + close(dma_buf_fd1);
> + close(dma_buf_fd2);
> + gem_close(fd, handle1);
> + gem_close(fd, handle2);
> +}
> +
> +static int
> +check_for_dma_buf_mmap(void)
> +{
> + int dma_buf_fd;
> + char *ptr;
> + uint32_t handle;
> + int ret = 1;
> +
> + handle = gem_create(fd, BO_SIZE);
> + dma_buf_fd = prime_handle_to_fd(fd, handle);
> + ptr = mmap(NULL, BO_SIZE, PROT_READ, MAP_SHARED, dma_buf_fd, 0);
> + if (ptr != MAP_FAILED)
> + ret = 0;
> + munmap(ptr, BO_SIZE);
> + gem_close(fd, handle);
> + close (dma_buf_fd);
> + return ret;
> +}
> +
> +igt_main
> +{
> + struct {
> + const char *name;
> + void (*fn)(void);
> + } tests[] = {
> + { "test_correct", test_correct },
> + { "test_map_unmap", test_map_unmap },
> + { "test_reprime", test_reprime },
> + { "test_forked", test_forked },
> + { "test_refcounting", test_refcounting },
> + { "test_dup", test_dup },
> + { "test_errors", test_errors },
> + { "test_aperture_limit", test_aperture_limit },
> + };
> + int i;
> +
> + igt_fixture {
> + fd = drm_open_any();
> + errno = 0;
> + }
> +
> + igt_skip_on((check_for_dma_buf_mmap() != 0));
> +
> + for (i = 0; i < ARRAY_SIZE(tests); i++) {
> + igt_subtest(tests[i].name)
> + tests[i].fn();
> + }
> +
> + igt_fixture
> + close(fd);
> +}
> --
> 2.1.0
>
--
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
next prev parent reply other threads:[~2015-08-13 7:04 UTC|newest]
Thread overview: 26+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-08-12 23:29 [PATCH v4] mmap on the dma-buf directly Tiago Vignatti
2015-08-12 23:29 ` [PATCH 1/4] drm: prime: Honour O_RDWR during prime-handle-to-fd Tiago Vignatti
2015-08-12 23:29 ` [PATCH 2/4] dma-buf: Add ioctls to allow userspace to flush Tiago Vignatti
2015-08-12 23:29 ` [PATCH 3/4] drm/i915: Implement end_cpu_access Tiago Vignatti
2015-08-23 12:14 ` Chris Wilson
2015-08-12 23:29 ` [PATCH 4/4] drm/i915: Use CPU mapping for userspace dma-buf mmap() Tiago Vignatti
2015-08-23 12:21 ` Chris Wilson
2015-08-12 23:29 ` [PATCH 1/7] prime_mmap: Add new test for calling mmap() on dma-buf fds Tiago Vignatti
2015-08-13 7:04 ` Daniel Vetter [this message]
2015-08-14 22:17 ` Tiago Vignatti
2015-08-25 7:46 ` Daniel Vetter
2015-08-12 23:29 ` [PATCH 2/7] prime_mmap: Fix a few misc stuff Tiago Vignatti
2015-08-12 23:29 ` [PATCH 3/7] prime_mmap: Add basic tests to write in a bo using CPU Tiago Vignatti
2015-08-13 7:01 ` Daniel Vetter
2015-08-13 14:26 ` Tiago Vignatti
2015-08-13 14:48 ` Daniel Vetter
2015-08-12 23:29 ` [PATCH 4/7] lib: Add gem_userptr helpers Tiago Vignatti
2015-08-23 12:10 ` Chris Wilson
2015-08-25 7:47 ` Daniel Vetter
2015-08-12 23:29 ` [PATCH 5/7] prime_mmap: Test for userptr mmap Tiago Vignatti
2015-08-23 12:17 ` Chris Wilson
2015-08-12 23:29 ` [PATCH 6/7] tests: Add kms_mmap_write_crc for cache coherency tests Tiago Vignatti
2015-08-12 23:29 ` [PATCH 7/7] tests/kms_mmap_write_crc: Demonstrate the need for end_cpu_access Tiago Vignatti
2015-08-13 11:09 ` [PATCH v4] mmap on the dma-buf directly Thomas Hellstrom
2015-08-13 14:09 ` Tiago Vignatti
2015-08-13 14:51 ` Daniel Vetter
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20150813070411.GV17734@phenom.ffwll.local \
--to=daniel@ffwll.ch \
--cc=daniel.thompson@linaro.org \
--cc=daniel.vetter@ffwll.ch \
--cc=dri-devel@lists.freedesktop.org \
--cc=intel-gfx@lists.freedesktop.org \
--cc=rob@linux.intel.com \
--cc=sumit.semwal@linaro.org \
--cc=tiago.vignatti@intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox