On Wed, Oct 25, 2017 at 05:28:39PM -0700, Antonio Argenziano wrote:
> This patch adds a context creation ioctl wrapper that returns the error
> for the caller to consume. Multiple tests that implemented this already,
> have been changed to use the new library function.
>
> v2:
> - Add gem_require_contexts() to check for contexts support (Chris)
>
> v3:
> - Add gem_has_contexts to check for contexts support and change
> gem_require_contexts to skip if contests support is not available.
> (Chris)
>
> Signed-off-by: Antonio Argenziano <antonio.argenziano@xxxxxxxxx>
>
> Cc: Chris Wilson <chris@xxxxxxxxxxxxxxxxxx>
> Cc: Michał Winiarski <michal.winiarski@xxxxxxxxx>
> ---
> benchmarks/gem_exec_ctx.c | 6 ++---
> benchmarks/gem_exec_trace.c | 4 +--
> lib/i915/gem_context.c | 62 +++++++++++++++++++++++++++++++++++++--------
> lib/i915/gem_context.h | 3 +++
> tests/gem_ctx_create.c | 10 ++++----
> tests/gem_ctx_switch.c | 13 ----------
> tests/gem_eio.c | 13 +---------
> tests/gem_exec_await.c | 14 ++--------
> tests/gem_exec_nop.c | 13 ----------
> tests/gem_exec_parallel.c | 15 +++--------
> tests/gem_exec_reuse.c | 13 ----------
> tests/gem_exec_whisper.c | 13 ----------
> 12 files changed, 71 insertions(+), 108 deletions(-)
>
> diff --git a/benchmarks/gem_exec_ctx.c b/benchmarks/gem_exec_ctx.c
> index 0eac04b0..a1c6e5d7 100644
> --- a/benchmarks/gem_exec_ctx.c
> +++ b/benchmarks/gem_exec_ctx.c
> @@ -64,7 +64,7 @@ static uint32_t batch(int fd)
> return handle;
> }
>
> -static uint32_t __gem_context_create(int fd)
> +static uint32_t __gem_context_create_local(int fd)
We only need this _local helper in negative parameters checks (in
tests/gem_ctx_create - and perhaps we can open-code those two ioctl calls?).
Both gem_exec_trace, gem_exec_ctx can use the helpers from lib.
> {
> struct drm_i915_gem_context_create create;
>
> @@ -101,7 +101,7 @@ static int loop(unsigned ring,
> execbuf.flags |= LOCAL_I915_EXEC_HANDLE_LUT;
> execbuf.flags |= LOCAL_I915_EXEC_NO_RELOC;
> if (mode != DEFAULT) {
> - execbuf.rsvd1 = __gem_context_create(fd);
> + execbuf.rsvd1 = __gem_context_create_local(fd);
> if (execbuf.rsvd1 == 0)
> return 77;
> }
> @@ -125,7 +125,7 @@ static int loop(unsigned ring,
> uint32_t ctx = 0;
>
> if (mode != DEFAULT && mode != NOP) {
> - execbuf.rsvd1 = __gem_context_create(fd);
> + execbuf.rsvd1 = __gem_context_create_local(fd);
> ctx = gem_context_create(fd);
> }
>
> diff --git a/benchmarks/gem_exec_trace.c b/benchmarks/gem_exec_trace.c
> index 12577649..2724ee92 100644
> --- a/benchmarks/gem_exec_trace.c
> +++ b/benchmarks/gem_exec_trace.c
> @@ -105,7 +105,7 @@ static double elapsed(const struct timespec *start, const struct timespec *end)
> return 1e3*(end->tv_sec - start->tv_sec) + 1e-6*(end->tv_nsec - start->tv_nsec);
> }
>
> -static uint32_t __gem_context_create(int fd)
> +static uint32_t __gem_context_create_local(int fd)
> {
> struct drm_i915_gem_context_create arg = {};
> drmIoctl(fd, DRM_IOCTL_I915_GEM_CONTEXT_CREATE, &arg);
> @@ -216,7 +216,7 @@ static double replay(const char *filename, long nop, long range)
> num_ctx = new_ctx;
> }
>
> - ctx[t->handle] = __gem_context_create(fd);
> + ctx[t->handle] = __gem_context_create_local(fd);
> break;
> }
> case DEL_CTX:
> diff --git a/lib/i915/gem_context.c b/lib/i915/gem_context.c
> index 6d9edf5e..a6c37051 100644
> --- a/lib/i915/gem_context.c
> +++ b/lib/i915/gem_context.c
> @@ -43,6 +43,22 @@
> * software features improving submission model (context priority).
> */
>
> +int __gem_context_create(int fd, uint32_t *ctx_id)
> +{
> + struct drm_i915_gem_context_create create;
> + int err = 0;
> +
> + memset(&create, 0, sizeof(create));
> + if (igt_ioctl(fd, DRM_IOCTL_I915_GEM_CONTEXT_CREATE, &create) == 0)
> + *ctx_id = create.ctx_id;
> + else
> + err = -errno;
> +
> + errno = 0;
> + return err;
> +}
> +
> +
Stray newline
> /**
> * gem_context_create:
> * @fd: open i915 drm file descriptor
> @@ -55,18 +71,44 @@
> */
> uint32_t gem_context_create(int fd)
> {
> - struct drm_i915_gem_context_create create;
> + uint32_t ctx_id;
> + gem_require_contexts(fd);
>
> - memset(&create, 0, sizeof(create));
> - if (igt_ioctl(fd, DRM_IOCTL_I915_GEM_CONTEXT_CREATE, &create)) {
> - int err = -errno;
> - igt_skip_on(err == -ENODEV || errno == -EINVAL);
> - igt_assert_eq(err, 0);
> - }
> - igt_assert(create.ctx_id != 0);
> - errno = 0;
> + igt_assert_eq(__gem_context_create(fd, &ctx_id), 0);
> + igt_assert(ctx_id != 0);
> +
> + return ctx_id;
> +}
>
> - return create.ctx_id;
> +/**
> + * gem_has_contexts:
> + * @fd: open i915 drm file descriptor
> + *
> + * Queries whether context creation is supported or not.
> + *
> + * Returns: Context creation availability.
> + */
> +bool gem_has_contexts(int fd)
> +{
> + uint32_t ctx_id = 0;
> +
> + __gem_context_create(fd, &ctx_id);
> + if (ctx_id)
> + gem_context_destroy(fd, ctx_id);
> +
> + return ctx_id;
> +}
> +
> +/**
> + * gem_require_contexts:
> + * @fd: open i915 drm file descriptor
> + *
> + * This helper will automatically skip the test on platforms where context
> + * support is not available.
> + */
> +void gem_require_contexts(int fd)
> +{
> + igt_require(gem_has_contexts(fd));
> }
Since we're exporting all of the functions, perhaps we can move
gem_request_contexts and gem_has_contexts, to keep create/destroy
pair next to each other?
>
> int __gem_context_destroy(int fd, uint32_t ctx_id)
> diff --git a/lib/i915/gem_context.h b/lib/i915/gem_context.h
> index a2339c4b..efff96b6 100644
> --- a/lib/i915/gem_context.h
> +++ b/lib/i915/gem_context.h
> @@ -24,6 +24,7 @@
> #ifndef GEM_CONTEXT_H
> #define GEM_CONTEXT_H
>
> +int __gem_context_create(int fd, uint32_t *ctx_id);
> uint32_t gem_context_create(int fd);
> void gem_context_destroy(int fd, uint32_t ctx_id);
> int __gem_context_destroy(int fd, uint32_t ctx_id);
Move __gem_context_create under gem_context_create to be consistent with
gem_context_destroy, but that's just my OCD talking.
> @@ -38,6 +39,8 @@ struct local_i915_gem_context_param {
> #define LOCAL_CONTEXT_PARAM_BANNABLE 0x5
> uint64_t value;
> };
> +bool gem_has_contexts(int fd);
> +void gem_require_contexts(int fd);
> void gem_context_require_bannable(int fd);
> void gem_context_require_param(int fd, uint64_t param);
> void gem_context_get_param(int fd, struct local_i915_gem_context_param *p);
> diff --git a/tests/gem_ctx_create.c b/tests/gem_ctx_create.c
> index ae0825a1..abe9a32e 100644
> --- a/tests/gem_ctx_create.c
> +++ b/tests/gem_ctx_create.c
> @@ -45,7 +45,7 @@ static unsigned all_nengine;
> static unsigned ppgtt_engines[16];
> static unsigned ppgtt_nengine;
>
> -static int __gem_context_create(int fd, struct drm_i915_gem_context_create *arg)
> +static int __gem_context_create_local(int fd, struct drm_i915_gem_context_create *arg)
> {
> int ret = 0;
> if (drmIoctl(fd, DRM_IOCTL_I915_GEM_CONTEXT_CREATE, arg))
> @@ -255,7 +255,7 @@ static void maximum(int fd, int ncpus, unsigned mode)
>
> err = -ENOMEM;
> if (avail_mem > (count + 1) * ctx_size)
> - err = __gem_context_create(fd, &create);
> + err = __gem_context_create_local(fd, &create);
We can also use the helper from lib here.
With that:
Reviewed-by: Michał Winiarski <michal.winiarski@xxxxxxxxx>
-Michał
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
