vulkan : improve ggml_vk_create_buffer error handling #9898
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
github-actions
bot
added
Vulkan
4 tasks
arthw
pushed a commit
to arthw/llama.cpp
that referenced
this pull request
Nov 15, 2024
arthw
pushed a commit
to arthw/llama.cpp
that referenced
this pull request
Nov 18, 2024
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This PR solves a segfault on Mali GPU. Sorry I can't give an easy reproduction since I can't reproduce it on my desktop GPU.
In
ggml_vk_create_buffer, whendevice->device.createBufferthrows an exception (e.g. out of memory ifsizeis too large), the program will crash. Note in this casebuf->sizeis already set butbuf->deviceis not set, the destructor ofvk_buffer_structwill cause trouble since its device is invalid.The PR solves the problem by setting
buf->sizeonly at the end of success path. Now when the context length is too long or model is too large, it will print a proper error message instead of segfault.