From: Tiago Vignatti <tiago.vignatti@intel.com>
To: dri-devel@lists.freedesktop.org
Cc: daniel.thompson@linaro.org, marcheu@google.com,
daniel.vetter@ffwll.ch, Rob Bradford <rob@linux.intel.com>,
thellstrom@vmware.com, jglisse@redhat.com, reveman@google.com
Subject: [PATCH igt v7 2/6] prime_mmap: Add new test for calling mmap() on dma-buf fds
Date: Tue, 22 Dec 2015 19:36:50 -0200 [thread overview]
Message-ID: <1450820214-12509-8-git-send-email-tiago.vignatti@intel.com> (raw)
In-Reply-To: <1450820214-12509-1-git-send-email-tiago.vignatti@intel.com>
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 checks that dup()ing the fd works
- test_userptr make sure it fails when mmaping due the lack of obj->base.filp
in a userptr.
- test_errors checks the error return values for failures
- test_aperture_limit tests multiple buffer creation at the gtt aperture
limit
v2 (Tiago): Removed pattern_check(), which was walking through a useless
iterator. Removed superfluous PROT_WRITE from gem_mmap, in test_correct().
Added binary file to .gitignore
v3 (Tiago): squash patch "prime_mmap: Test for userptr mmap" into this one.
v4 (Tiago): use synchronized userptr for testing. Add test for buffer
overlapping.
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 | 417 +++++++++++++++++++++++++++++++++++++++++++++++++
2 files changed, 418 insertions(+)
create mode 100644 tests/prime_mmap.c
diff --git a/tests/Makefile.sources b/tests/Makefile.sources
index d594038..75f3cb0 100644
--- a/tests/Makefile.sources
+++ b/tests/Makefile.sources
@@ -96,6 +96,7 @@ 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..95304a9
--- /dev/null
+++ b/tests/prime_mmap.c
@@ -0,0 +1,417 @@
+/*
+ * 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 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__gtt(fd, handle, BO_SIZE, PROT_READ);
+ 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(memcmp(ptr2, pattern, sizeof(pattern)) == 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(memcmp(ptr, pattern, sizeof(pattern)) == 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(memcmp(ptr, pattern, sizeof(pattern)) == 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(memcmp(ptr, pattern, sizeof(pattern)) == 0);
+
+ close (dma_buf_fd);
+ igt_assert(memcmp(ptr, pattern, sizeof(pattern)) == 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(memcmp(ptr, pattern, sizeof(pattern)) == 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(memcmp(ptr, pattern, sizeof(pattern)) == 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(memcmp(ptr, pattern, sizeof(pattern)) == 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(memcmp(ptr, pattern, sizeof(pattern)) == 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);
+ if (ret)
+ ret = errno;
+ *fd_out = args.fd;
+
+ return ret;
+}
+
+/* test for mmap(dma_buf_export(userptr)) */
+static void
+test_userptr(void)
+{
+ int ret, dma_buf_fd;
+ void *ptr;
+ uint32_t handle;
+
+ /* create userptr bo */
+ ret = posix_memalign(&ptr, 4096, BO_SIZE);
+ igt_assert_eq(ret, 0);
+
+ /* we are not allowed to export unsynchronized userptr. Just create a normal
+ * one */
+ gem_userptr(fd, (uint32_t *)ptr, BO_SIZE, 0, 0, &handle);
+
+ /* export userptr */
+ ret = prime_handle_to_fd_no_assert(handle, &dma_buf_fd);
+ if (ret) {
+ igt_assert(ret == EINVAL || ret == ENODEV);
+ goto free_userptr;
+ } else {
+ igt_assert_eq(ret, 0);
+ igt_assert_lte(0, dma_buf_fd);
+ }
+
+ /* a userptr doesn't have the obj->base.filp, but can be exported via
+ * dma-buf, so make sure it fails here */
+ ptr = mmap(NULL, BO_SIZE, PROT_READ, MAP_SHARED, dma_buf_fd, 0);
+ igt_assert(ptr == MAP_FAILED && errno == ENODEV);
+free_userptr:
+ gem_close(fd, handle);
+ close(dma_buf_fd);
+}
+
+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);
+
+ /* Overlapping the end of the buffer */
+ handle = gem_create(fd, 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, BO_SIZE / 2);
+ igt_assert(ptr == MAP_FAILED && errno == EINVAL);
+ errno = 0;
+ close(dma_buf_fd);
+ gem_close(fd, handle);
+}
+
+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(memcmp(ptr1, pattern, sizeof(pattern)) == 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(memcmp(ptr2, pattern, sizeof(pattern)) == 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_userptr", test_userptr },
+ { "test_errors", test_errors },
+ { "test_aperture_limit", test_aperture_limit },
+ };
+ int i;
+
+ igt_fixture {
+ fd = drm_open_driver(DRIVER_INTEL);
+ 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);
+}
--
2.1.4
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/dri-devel
next prev parent reply other threads:[~2015-12-22 21:37 UTC|newest]
Thread overview: 43+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-12-22 21:36 Direct userspace dma-buf mmap (v7) Tiago Vignatti
2015-12-22 21:36 ` [PATCH v7 1/5] drm: prime: Honour O_RDWR during prime-handle-to-fd Tiago Vignatti
2015-12-22 21:36 ` [PATCH v7 2/5] dma-buf: Remove range-based flush Tiago Vignatti
2015-12-22 21:36 ` [PATCH v7 3/5] dma-buf: Add ioctls to allow userspace to flush Tiago Vignatti
2016-02-09 9:26 ` David Herrmann
2016-02-09 10:20 ` Daniel Vetter
2016-02-09 10:52 ` Daniel Vetter
2016-02-11 17:54 ` Tiago Vignatti
2016-02-11 18:00 ` Alex Deucher
2016-02-11 18:08 ` David Herrmann
2016-02-11 18:08 ` Ville Syrjälä
2016-02-11 18:19 ` David Herrmann
2016-02-11 19:10 ` Ville Syrjälä
2016-02-11 22:04 ` [PATCH v9] " Tiago Vignatti
2016-02-12 14:50 ` David Herrmann
2016-02-12 15:02 ` Daniel Vetter
2016-02-25 18:01 ` Chris Wilson
2016-02-29 14:54 ` Daniel Vetter
2016-02-29 15:02 ` Chris Wilson
2016-03-05 9:34 ` Daniel Vetter
2016-03-14 20:21 ` Tiago Vignatti
2016-03-15 8:51 ` Chris Wilson
2016-03-17 18:18 ` [PATCH] prime_mmap_coherency: Add return error tests for prime sync ioctl Tiago Vignatti
2016-03-17 21:01 ` Chris Wilson
2016-03-17 21:15 ` Tiago Vignatti
2016-03-17 21:18 ` [PATCH v2] " Tiago Vignatti
2016-03-18 9:44 ` Chris Wilson
2016-03-18 9:53 ` [Intel-gfx] " Chris Wilson
2016-03-18 18:08 ` [PATCH v3] " Tiago Vignatti
2016-03-18 18:11 ` Daniel Vetter
2016-03-18 18:17 ` Tiago Vignatti
2016-03-18 20:43 ` Chris Wilson
2015-12-22 21:36 ` [PATCH v7 4/5] drm/i915: Implement end_cpu_access Tiago Vignatti
2015-12-22 21:36 ` [PATCH v7 5/5] drm/i915: Use CPU mapping for userspace dma-buf mmap() Tiago Vignatti
2015-12-22 21:36 ` [PATCH igt v7 1/6] lib: Add gem_userptr and __gem_userptr helpers Tiago Vignatti
2015-12-22 21:36 ` Tiago Vignatti [this message]
2015-12-22 21:36 ` [PATCH igt v7 3/6] prime_mmap: Add basic tests to write in a bo using CPU Tiago Vignatti
2015-12-22 21:36 ` [PATCH igt v7 4/6] lib: Add prime_sync_start and prime_sync_end helpers Tiago Vignatti
2015-12-22 21:36 ` [PATCH igt v7 5/6] tests: Add kms_mmap_write_crc for cache coherency tests Tiago Vignatti
2015-12-22 21:36 ` [PATCH igt v7 6/6] tests: Add prime_mmap_coherency " Tiago Vignatti
2016-02-04 20:55 ` Direct userspace dma-buf mmap (v7) Stéphane Marchesin
2016-02-05 13:53 ` Tiago Vignatti
2016-02-09 8:47 ` 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=1450820214-12509-8-git-send-email-tiago.vignatti@intel.com \
--to=tiago.vignatti@intel.com \
--cc=daniel.thompson@linaro.org \
--cc=daniel.vetter@ffwll.ch \
--cc=dri-devel@lists.freedesktop.org \
--cc=jglisse@redhat.com \
--cc=marcheu@google.com \
--cc=reveman@google.com \
--cc=rob@linux.intel.com \
--cc=thellstrom@vmware.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;
as well as URLs for NNTP newsgroup(s).