[PATCH] [RFC][1/17]TILE-Gx: recognize tilegx triple

[Jiong Wang]

于 2013/3/11 21:30, Joerg Sonnenberger 写道:
> On Sun, Mar 10, 2013 at 11:56:31PM -0700, Jiong WANG wrote:
>>    this is the first patch which teach llvm to recognize tilegx triple.
> Please add a case in the --enable-jit switch to explicitly disable it.
>
>> @@ -780,13 +782,13 @@
>>   AC_ARG_ENABLE([targets],AS_HELP_STRING([--enable-targets],
>>       [Build specific host targets: all or target1,target2,... Valid targets are:
>>        host, x86, x86_64, sparc, powerpc, arm, aarch64, mips, hexagon,
>> -     xcore, msp430, nvptx, and cpp (default=all)]),,
>> +     xcore, msp430, nvptx, tilegx, and cpp (default=all)]),,
> -> "all, but tilegx"
>
>>       enableval=all)
>>   if test "$enableval" = host-only ; then
>>     enableval=host
>>   fi
>>   case "$enableval" in
>> -  all) TARGETS_TO_BUILD="X86 Sparc PowerPC AArch64 ARM Mips XCore MSP430 CppBackend MBlaze NVPTX Hexagon" ;;
>> +  all) TARGETS_TO_BUILD="X86 Sparc PowerPC AArch64 ARM Mips XCore MSP430 CppBackend MBlaze NVPTX Hexagon Tile" ;;
>>     *)for a_target in `echo $enableval|sed -e 's/,/ /g' ` ; do
>>         case "$a_target" in
>>           x86)      TARGETS_TO_BUILD="X86 $TARGETS_TO_BUILD" ;;
> Do not include it here, so that it isn't enabled by default.
> Otherwise, LGTM.
Joerg,

thanks for your time to review.

I know the maturity of a new target is very important to be accepted, 
then what's the requirement for a target to be enabled by default?

tilegx actually doing well on regression tests and test-suites. I 
checked http://llvm.org/perf/db_default/v4/nts/recent_activity, tilegx 
has fewer unexpected failures than powerpc and arm.

---
Regards,
Jiong
Tilera Corporation
>
> Joerg
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits