[igt-dev] [PATCH i-g-t 3/3] lib/intel_allocator: Add support for regions of allocation
Zbigniew Kempczyński
zbigniew.kempczynski at intel.com
Fri Jun 11 08:57:35 UTC 2021
On Wed, Jun 09, 2021 at 03:15:22PM +0200, Andrzej Turko wrote:
> Enable creating allocators managing only a given interval
> of available gtt for all implementations, not just the
> simple allocator. Additinally, set a universal lower
> bound on alignment to 4096.
>
> Signed-off-by: Andrzej Turko <andrzej.turko at linux.intel.com>
> Cc: Zbigniew Kempczyński <zbigniew.kempczynski at intel.com>
> ---
> lib/intel_allocator.c | 14 +++++++++-----
> lib/intel_allocator_random.c | 30 +++++++++++++++++++-----------
> lib/intel_allocator_reloc.c | 18 ++++++++++++------
> lib/intel_allocator_simple.c | 1 -
> 4 files changed, 40 insertions(+), 23 deletions(-)
>
> diff --git a/lib/intel_allocator.c b/lib/intel_allocator.c
> index a78a661b1..82875224d 100644
> --- a/lib/intel_allocator.c
> +++ b/lib/intel_allocator.c
> @@ -64,8 +64,10 @@ struct handle_entry {
> struct allocator *al;
> };
>
> -struct intel_allocator *intel_allocator_reloc_create(int fd, uint64_t end);
> -struct intel_allocator *intel_allocator_random_create(int fd, uint64_t end);
> +struct intel_allocator *
> +intel_allocator_reloc_create(int fd, uint64_t start, uint64_t end);
> +struct intel_allocator *
> +intel_allocator_random_create(int fd, uint64_t start, uint64_t end);
> struct intel_allocator *
> intel_allocator_simple_create(int fd, uint64_t start, uint64_t end,
> enum allocator_strategy strategy);
> @@ -291,10 +293,10 @@ static struct intel_allocator *intel_allocator_create(int fd,
> "We cannot use NONE allocator\n");
> break;
> case INTEL_ALLOCATOR_RELOC:
> - ial = intel_allocator_reloc_create(fd, end);
> + ial = intel_allocator_reloc_create(fd, start, end);
> break;
> case INTEL_ALLOCATOR_RANDOM:
> - ial = intel_allocator_random_create(fd, end);
> + ial = intel_allocator_random_create(fd, start, end);
> break;
> case INTEL_ALLOCATOR_SIMPLE:
> ial = intel_allocator_simple_create(fd, start, end,
> @@ -1078,10 +1080,12 @@ uint64_t __intel_allocator_alloc(uint64_t allocator_handle, uint32_t handle,
> .allocator_handle = allocator_handle,
> .alloc.handle = handle,
> .alloc.size = size,
> - .alloc.alignment = alignment,
> .alloc.strategy = strategy };
> struct alloc_resp resp;
>
> + igt_assert((alignment & (alignment-1)) == 0);
> + req.alloc.alignment = max(alignment, 1 << 12);
> +
> igt_assert(handle_request(&req, &resp) == 0);
> igt_assert(resp.response_type == RESP_ALLOC);
>
> diff --git a/lib/intel_allocator_random.c b/lib/intel_allocator_random.c
> index 85ac2cf4e..1d2969618 100644
> --- a/lib/intel_allocator_random.c
> +++ b/lib/intel_allocator_random.c
> @@ -10,12 +10,12 @@
> #include "igt_rand.h"
> #include "intel_allocator.h"
>
> -struct intel_allocator *intel_allocator_random_create(int fd, uint64_t end);
> +struct intel_allocator *
> +intel_allocator_random_create(int fd, uint64_t start, uint64_t end);
>
> struct intel_allocator_random {
> - uint64_t bias;
> uint32_t prng;
> - uint64_t address_mask;
> + uint64_t size_mask;
> uint64_t start;
> uint64_t end;
>
> @@ -23,6 +23,7 @@ struct intel_allocator_random {
> uint64_t allocated_objects;
> };
>
> +/* Keep the low 256k clear, for negative deltas */
> #define BIAS 256 << 10
>
#define RETRIES 8
> static void intel_allocator_random_get_address_range(struct intel_allocator *ial,
> @@ -45,16 +46,22 @@ static uint64_t intel_allocator_random_alloc(struct intel_allocator *ial,
> {
> struct intel_allocator_random *ialr = ial->priv;
> uint64_t offset;
> + int cnt = 0;
int cnt = RETRIES;
>
> (void) handle;
> (void) strategy;
>
> /* randomize the address, we try to avoid relocations */
> do {
> + /* the object won't fit in the address range */
> + if (cnt > 3) return ALLOC_INVALID_ADDRESS;
> +
> offset = hars_petruska_f54_1_random64(&ialr->prng);
> - offset += ialr->bias; /* Keep the low 256k clear, for negative deltas */
> - offset &= ialr->address_mask;
> - offset &= ~(alignment - 1);
> + if (cnt++ == 3) offset = 0;
> +
> + offset &= ialr->size_mask;
> + offset += ialr->start;
> + offset = ALIGN(offset, alignment);
> } while (offset + size > ialr->end);
Maybe:
/* We won't fit at all */
if (size > ialr->end - ialr->start)
return ALLOC_INVALID_ADDRESS;
do {
offset = hars_petruska_f54_1_random64(&ialr->prng);
offset &= ialr->size_mask;
offset += ialr->start;
offset = ALIGN(offset, alignment);
} while (offset + size > ialr->end && cnt--);
if (!cnt)
return ALLOC_INVALID_ADDRESS;
will be more readable?
>
> ialr->allocated_objects++;
> @@ -145,7 +152,8 @@ static bool intel_allocator_random_is_empty(struct intel_allocator *ial)
> return !ialr->allocated_objects;
> }
>
> -struct intel_allocator *intel_allocator_random_create(int fd, uint64_t end)
> +struct intel_allocator *
> +intel_allocator_random_create(int fd, uint64_t start, uint64_t end)
> {
> struct intel_allocator *ial;
> struct intel_allocator_random *ialr;
> @@ -170,10 +178,10 @@ struct intel_allocator *intel_allocator_random_create(int fd, uint64_t end)
> igt_assert(ial->priv);
> ialr->prng = (uint32_t) to_user_pointer(ial);
>
> - igt_assert(BIAS < end);
> - ialr->address_mask = (1ULL << (igt_fls(end) - 1)) - 1;
> - ialr->bias = BIAS;
> - ialr->start = ialr->bias;
> + start = max(start, BIAS);
> + igt_assert(start < end);
> + ialr->size_mask = (1ULL << (igt_fls(end - start) - 1)) - 1;
> + ialr->start = start;
> ialr->end = end;
No, mask will be invalid in this case.
--
Zbigniew
>
> ialr->allocated_objects = 0;
> diff --git a/lib/intel_allocator_reloc.c b/lib/intel_allocator_reloc.c
> index 703d054c6..24b2aeebc 100644
> --- a/lib/intel_allocator_reloc.c
> +++ b/lib/intel_allocator_reloc.c
> @@ -10,10 +10,10 @@
> #include "igt_rand.h"
> #include "intel_allocator.h"
>
> -struct intel_allocator *intel_allocator_reloc_create(int fd, uint64_t end);
> +struct intel_allocator *
> +intel_allocator_reloc_create(int fd, uint64_t start, uint64_t end);
>
> struct intel_allocator_reloc {
> - uint64_t bias;
> uint32_t prng;
> uint64_t start;
> uint64_t end;
> @@ -23,6 +23,7 @@ struct intel_allocator_reloc {
> uint64_t allocated_objects;
> };
>
> +/* Keep the low 256k clear, for negative deltas */
> #define BIAS 256 << 10
>
> static void intel_allocator_reloc_get_address_range(struct intel_allocator *ial,
> @@ -49,13 +50,16 @@ static uint64_t intel_allocator_reloc_alloc(struct intel_allocator *ial,
> (void) handle;
> (void) strategy;
>
> - alignment = max(alignment, 4096);
> aligned_offset = ALIGN(ialr->offset, alignment);
>
> /* Check we won't exceed end */
> if (aligned_offset + size > ialr->end)
> aligned_offset = ALIGN(ialr->start, alignment);
>
> + /* Check that the object fits in the address range */
> + if (aligned_offset + size > ialr->end)
> + return ALLOC_INVALID_ADDRESS;
> +
> offset = aligned_offset;
> ialr->offset = offset + size;
> ialr->allocated_objects++;
> @@ -146,7 +150,8 @@ static bool intel_allocator_reloc_is_empty(struct intel_allocator *ial)
> return !ialr->allocated_objects;
> }
>
> -struct intel_allocator *intel_allocator_reloc_create(int fd, uint64_t end)
> +struct intel_allocator *
> +intel_allocator_reloc_create(int fd, uint64_t start, uint64_t end)
> {
> struct intel_allocator *ial;
> struct intel_allocator_reloc *ialr;
> @@ -171,8 +176,9 @@ struct intel_allocator *intel_allocator_reloc_create(int fd, uint64_t end)
> igt_assert(ial->priv);
> ialr->prng = (uint32_t) to_user_pointer(ial);
>
> - ialr->bias = ialr->offset = BIAS;
> - ialr->start = ialr->bias;
> + start = max(start, BIAS);
> + igt_assert(start < end);
> + ialr->offset = ialr->start = start;
> ialr->end = end;
>
> ialr->allocated_objects = 0;
> diff --git a/lib/intel_allocator_simple.c b/lib/intel_allocator_simple.c
> index 6022e832b..0e6763964 100644
> --- a/lib/intel_allocator_simple.c
> +++ b/lib/intel_allocator_simple.c
> @@ -419,7 +419,6 @@ static uint64_t intel_allocator_simple_alloc(struct intel_allocator *ial,
> ials = (struct intel_allocator_simple *) ial->priv;
> igt_assert(ials);
> igt_assert(handle);
> - alignment = alignment > 0 ? alignment : 1;
>
> rec = igt_map_search(ials->objects, &handle);
> if (rec) {
> --
> 2.25.1
>
More information about the igt-dev
mailing list