vulkan: remove the need for the dryrun (llama/16826)

* vulkan: remove the need for the dryrun

Allocate pipelines and descriptor sets when requested.

Reallocate the prealloc buffers when needed, and flush any pending work
before reallocating.

For rms_partials and total_mul_mat_bytes, use the sizes computed the last time
the graph was executed.

* remove dryrun parameters
This commit is contained in:
Jeff Bolz
2025-11-09 23:38:03 +02:00
committed by Georgi Gerganov
parent 997fdde0c4
commit 1672d41ab0
File diff suppressed because it is too large Load Diff