Fix overlaps.sql test fail on 32 bit Debian due to gcc bug 323 #62
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
On some 32 bit platforms, there is a gcc bug that makes floating point calculations and comparisons unstable (see the link below). The problem originates in FPU 80 bits registers where double values are not truncated to 64 bit values. When gcc compiles some code with enabled optimizations, the intermediate results may be kept in the FPU registers without truncation to 64 bit values. Extra bits may produce unstable results when comparing the numbers. The generic solution is to save the intermediate results in the memory where the values are truncated to 64 bit values. It affects the performance but makes the tests stable on all platforms. PGSPHERE_FLOAT_STORE macro enables storing of intermediate results for FPxx operations in the memory. It is enabled by default for 32 bit platforms. It can be explicitly enabled or disabled in CFLAGS. To enable it for all code the gcc option -ffloat-store may be used as well. Link to gcc bug: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=323
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
On some 32 bit platforms, there is a gcc bug that makes floating point calculations and comparisons unstable (see the link below). The problem originates in FPU 80 bits registers where double values are not truncated to 64 bit values. When gcc compiles some code with enabled optimizations, the intermediate results may be kept in the FPU registers without truncation to 64 bit values. Extra bits may produce unstable results when comparing the numbers.
The generic solution is to save the intermediate results in the memory where the values are truncated to 64 bit values. It affects the performance but makes the tests stable on all platforms.
PGSPHERE_FLOAT_STORE macro enables storing of intermediate results for FPxx operations in the memory. It is enabled by default for 32 bit platforms. It can be explicitly enabled or disabled in CFLAGS. To enable it for all code the gcc option -ffloat-store may be used as well.
Link to gcc bug: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=323