diff options
| author | Dave Bartolomeo <dave_bartolomeo@apple.com> | 2025-09-17 12:25:51 -0400 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2025-09-17 09:25:51 -0700 |
| commit | 48f00e81338e637b8210d3320527d4f389d5343b (patch) | |
| tree | 4970b82038742be9cfe15713e6c0227bc6d0d6d8 /llvm/lib/CodeGen/LiveDebugVariables.cpp | |
| parent | d90a313b5e90771c0956b5fdd2814bc93bf414de (diff) | |
| download | llvm-48f00e81338e637b8210d3320527d4f389d5343b.zip llvm-48f00e81338e637b8210d3320527d4f389d5343b.tar.gz llvm-48f00e81338e637b8210d3320527d4f389d5343b.tar.bz2 | |
[analyzer] Prevent triplicate warnings for `sarif-html` (#158112)
When `-analyzer-output=sarif-html` is specified, the analyzer was
reporting each warning to the console three times. This is because the
code to create the diagnostic consumers for `sarif-html` was calling the
code for `sarif` and `html` separately, each of which also creates its
own console text consumer. Then the `sarif-html` code itself created a
third.
The fix is to factor out the creation of the SARIF and HTML consumers
from the text consumers, so `sarif-html` just calls the code to create
the SARIF and HTML consumers without the text consumers.
The same fix applies for `plist-html`.
I've updated one of the SARIF tests to specify `sarif-html`. This test
would fail in the regular `-verify` validation due to the triplicated
warnings, but now passes with my fix.
Fixes #158103
rdar://160383710
Diffstat (limited to 'llvm/lib/CodeGen/LiveDebugVariables.cpp')
0 files changed, 0 insertions, 0 deletions