This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [patch] : Add / extend documentation and inline functions


Hi Chris,

> Adds some documentation to stl_algo.h for the new functions, and also
> improves some parts of the old documentation. I didn't bother fully
> documenting things about internal functions I sort were obvious
> (first, last are iterators, comp is a predicate / ordering, etc), but
> I did add some slightly subtle preconditions to the partition
> functions while I remember them :)

Thanks.

> This also inlines a large bunch of trivial functions, that I really
> should have marked inline before.

Unfortunately, this part of the patch "causes" a regression (x86-linux):

sort.exe:
/home/paolo/Gcc/cvs-dirs/libstdcxx_so_7-branch/gcc/libstdc++-v3/testsuite/25_algorithms
/sort/sort.cc:139: void test04(): Assertion `std::equal(s2,
partial_sort_copy(s1, s1 + N, s2, s2
+ 2*N), A)' failed.
FAIL: 25_algorithms/sort/sort.cc execution test

I'm sure it's a miscompilation (what else?) still, adding the patch
without providing some sort of feedback to the compiler people seems
*very* risky: if we are unlucky and the problem is not known or subsumed
by a different one, either merging the changes in v7 will become
impossible or will cause a regression in mainline much later. Can you
please investigate the issue in some detail?

Thanks,
Paolo.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]