Fix GCC 5.x issues so that it can be used in Windows

[Adrian Pop]

There are some tests failing when using GCC 5.x:
​https://test.openmodelica.org/hudson/job/OpenModelica_TEST_GCC_5/lastCompletedBuild/testReport/

These should be fixed before we update OMDev to msys2 which uses GCC 5.2.

[Adrian Pop]

Update: seems that if you compiled Matching.mo with -O1 it works fine.

[Adrian Pop]

So is one of these that needs to be disabled (or a combination) for Matching.mo:

-fthread-jumps -falign-functions  -falign-jumps -falign-loops 
-falign-labels -fcaller-saves -fcrossjumping -fcse-follow-jumps 
-fcse-skip-blocks -fdelete-null-pointer-checks 
-fdevirtualize -fdevirtualize-speculatively -fexpensive-optimizations 
-fgcse  -fgcse-lm -fhoist-adjacent-loads -finline-small-functions 
-findirect-inlining -fipa-cp -fipa-cp-alignment 
-fipa-sra -fipa-icf -fisolate-erroneous-paths-dereference 
-flra-remat -foptimize-sibling-calls -foptimize-strlen 
-fpartial-inlining -fpeephole2 -freorder-blocks -freorder-blocks-and-partition 
-freorder-functions -frerun-cse-after-loop -fsched-interblock -fsched-spec 
-fschedule-insns -fschedule-insns2 -fstrict-aliasing -fstrict-overflow 
-ftree-builtin-call-dce -ftree-switch-conversion -ftree-tail-merge 
-ftree-pre -ftree-vrp -fipa-ra

Anybody wants to hazard a guess? :)

[Adrian Pop]

The diffs between GCC 4.9.3 and GCC 5.2.0 with regards to -O2 optimizations are:

.txt

@@ -11 +11 @@
 -fhoist-adjacent-loads 
 -finline-small-functions 
 -findirect-inlining 
+-fipa-cp 
+-fipa-cp-alignment 
 -fipa-sra 
+-fipa-icf 
 -fisolate-erroneous-paths-dereference 
+-flra-remat 
 -foptimize-sibling-calls 
+-foptimize-strlen 
 -fpartial-inlining 
 -fpeephole2 
 -freorder-blocks  -freorder-functions 
+-freorder-blocks-and-partition 
 -frerun-cse-after-loop  
 -fsched-interblock  -fsched-spec 
 -fschedule-insns  -fschedule-insns2 
 -fstrict-aliasing -fstrict-overflow 
+-ftree-builtin-call-dce 
 -ftree-switch-conversion -ftree-tail-merge 
 -ftree-pre 
 -ftree-vrp
+-fipa-ra

So maybe one of the new ones are buggy :)

[Martin Sjölund]

If it's only Matching.mo, it is possible to script it. I could try in the afternoon when I am less busy, if you'd like me to.

[Adrian Pop]

I have tried the omc with Matching.mo compiled with -O1 just on one model: testsuite/simulation/modelica/built_in_functions/DummyDerMatching.mo so I'm not 100% sure it will work on all the affected ones.

[Martin Sjölund]

I've gone through all of those optimization, and -O1 -f$opt works for all of them. It must be a combination. Is this list of -O2 optimizations a complete list of GCC 5.2.0 -O2 optimization? Because -O1 -fALLOFTHEM... also works.

[Adrian Pop]

Yes, I've copied them from here and did the diff:

GCC 5.2.0:
​https://gcc.gnu.org/onlinedocs/gcc-5.2.0/gcc/Optimize-Options.html#Optimize-Options

GCC 4.9.3:
​https://gcc.gnu.org/onlinedocs/gcc-4.9.3/gcc/Optimize-Options.html#Optimize-Options

[Martin Sjölund]

Partial fix in 97de296: It was an -O1 flag that was marking function calls pure that was the culprit. GCC 5.1 will still be broken, but 5.2 seems to work (for C code, the tests use g++-4.9 for the cppruntime since boost breaks in g++-5.2).

[Martin Sjölund]

Moved to new milestone 1.9.4

[Martin Sjölund]

Milestone pushed to 1.9.5

[Martin Sjölund]

Milestone renamed

[Martin Sjölund]

Ticket retargeted after milestone closed

[Martin Sjölund]

Milestone moved to 1.12.0 due to 1.11.0 already being released.

[Francesco Casella]

The milestone of this ticket has been reassigned to "Future".

If you think the issue is still valid and relevant for you, please select milestone 1.13.0 for back-end, code generation and run-time issues, or 2.0.0 for front-end issues.

If you are aware that the problem is no longer present, please select the milestone corresponding to the version of OMC you used to check that, and set the status to "worksforme".

In both cases, a short informative comment would be welcome.