Skip to content

Fix coverity issues #30082

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 3 commits into
base: master
Choose a base branch
from

Conversation

itikhono
Copy link
Contributor

Details:

Fixed "dereference before null check" issues

Tickets:

@itikhono itikhono requested review from a team as code owners April 11, 2025 08:39
@github-actions github-actions bot added category: transformations OpenVINO Runtime library - Transformations category: LP transformations OpenVINO Low Precision transformations labels Apr 11, 2025
@itikhono itikhono requested a review from v-Golubev April 11, 2025 08:39
…ugh_dequantization.cpp

Co-authored-by: Andrii Staikov <andrii.staikov@intel.com>
@itikhono itikhono requested a review from CuriousPanCake April 11, 2025 08:43
Comment on lines +145 to 148
auto child = reshape_target_inputs.begin()->get_node();
if (ov::is_type<opset1::GroupConvolution>(child)) {
return false;
}
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Is there any reason why this is not a part of the pattern?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

not so sure, but probably because we check that GroupConvolution is the first child ( .begin() )
if we add this subgraph "reshape -> group_conv" to the pattern, we will have to check that group_conv is the first child anyway

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Actually if any of the consumers is GroupConvolution, we shouldn't perform this transformation.
Such dequantization subgraphs can be shared between several nodes with the same type in some rare cases (however, I haven't seen such cases with GroupConvolution), so we could stop this transformation when all the consumers are GroupConvs. But theoretically there may be some ShapeOf subgraphs starting from this reshape, so maybe "all_of" check is too strict

@itikhono itikhono requested a review from CuriousPanCake April 11, 2025 12:33
@itikhono itikhono enabled auto-merge April 11, 2025 12:56
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
category: LP transformations OpenVINO Low Precision transformations category: transformations OpenVINO Runtime library - Transformations
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants