Török Edwin
2007-Nov-15 19:24 UTC
[LLVMdev] Are 'make check' failures of interest in llvm-gcc?
Hi, I have run 'make check' in llvm-gcc 4.0 (svn version). I got some failures, mostly gcc.apple tests fail (and a gcc.c-torture test, namely 20000804-1.c) Are these tests of any interest to llvm-gcc itself, or they are problems with "upstream" ? (gcc 4.0). If the former, should I submit bugreports about them? Thanks, --Edwin
Dale Johannesen
2007-Nov-15 19:59 UTC
[LLVMdev] Are 'make check' failures of interest in llvm-gcc?
On Nov 15, 2007, at 11:24 AM, Török Edwin wrote:> Hi, > > I have run 'make check' in llvm-gcc 4.0 (svn version). > I got some failures, mostly gcc.apple tests fail (and a gcc.c-torture > test, namely 20000804-1.c) > > Are these tests of any interest to llvm-gcc itself, or they are > problems > with "upstream" ? (gcc 4.0). > If the former, should I submit bugreports about them?Could be either. I'm running this suite and seeing a lot more failures than you're reporting; many builtin optimizations are not done, many expected warnings do not match gcc syntax, etc. There are also some real bugs, but all the ones I know of are related to rarely used features, mostly gcc extensions. It would be more useful to do this in 4.2 IMO.
Török Edwin
2007-Nov-15 20:03 UTC
[LLVMdev] Are 'make check' failures of interest in llvm-gcc?
Dale Johannesen wrote:> On Nov 15, 2007, at 11:24 AM, Török Edwin wrote: > > Could be either. I'm running this suite and seeing a lot more > failures than > you're reporting;I didn't wait for all the tests to finish, just seen that there are many...> many builtin optimizations are not done, many expected > warnings do not match gcc syntax, etc. There are also some real bugs, > but all the ones I know of are related to rarely used features, > mostly gcc > extensions. > > It would be more useful to do this in 4.2 IMO. >Ok. --Edwin