Hi Richard, Ramana. FYI, trunk currently fails to bootstrap on ARM:
../../../gcc-4.7~/libcpp/expr.c: In function 'cpp_num_sign_extend': ../../../gcc-4.7~/libcpp/expr.c:1321:1: error: could not split insn (insn:TI 53 49 92 6 (set (reg:DI 10 sl [orig:151 num$high ] [151]) (ior:DI (not:DI (reg:DI 2 r2 [171])) (reg:DI 10 sl [orig:151 num$high ] [151]))) ../../../gcc-4.7~/libcpp/expr.c:1310 869 {orndi3_neon} (nil)) ../../../gcc-4.7~/libcpp/expr.c:1321:1: internal compiler error: in final_scan_insn, at final.c:2723
See: http://builds.linaro.org/toolchain/gcc-4.7~svn174320/logs/armv7l-maverick-cb...
-- Michael
On 29/05/11 22:17, Michael Hope wrote:
Hi Richard, Ramana. FYI, trunk currently fails to bootstrap on ARM:
../../../gcc-4.7~/libcpp/expr.c: In function 'cpp_num_sign_extend': ../../../gcc-4.7~/libcpp/expr.c:1321:1: error: could not split insn (insn:TI 53 49 92 6 (set (reg:DI 10 sl [orig:151 num$high ] [151]) (ior:DI (not:DI (reg:DI 2 r2 [171])) (reg:DI 10 sl [orig:151 num$high ] [151]))) ../../../gcc-4.7~/libcpp/expr.c:1310 869 {orndi3_neon} (nil)) ../../../gcc-4.7~/libcpp/expr.c:1321:1: internal compiler error: in final_scan_insn, at final.c:2723
I think this is because of my latest patch to canonicalize the orn patterns in neon.md. Let me have a look today.
Ramana
See: http://builds.linaro.org/toolchain/gcc-4.7~svn174320/logs/armv7l-maverick-cb...
-- Michael
linaro-toolchain mailing list linaro-toolchain@lists.linaro.org http://lists.linaro.org/mailman/listinfo/linaro-toolchain
On 29/05/11 22:17, Michael Hope wrote:
Hi Richard, Ramana. FYI, trunk currently fails to bootstrap on ARM:
../../../gcc-4.7~/libcpp/expr.c: In function 'cpp_num_sign_extend': ../../../gcc-4.7~/libcpp/expr.c:1321:1: error: could not split insn (insn:TI 53 49 92 6 (set (reg:DI 10 sl [orig:151 num$high ] [151]) (ior:DI (not:DI (reg:DI 2 r2 [171])) (reg:DI 10 sl [orig:151 num$high ] [151]))) ../../../gcc-4.7~/libcpp/expr.c:1310 869 {orndi3_neon} (nil)) ../../../gcc-4.7~/libcpp/expr.c:1321:1: internal compiler error: in final_scan_insn, at final.c:272
This is a patch I'm testing tonight for this - It's passed bootstrap in ARM and Thumb states beyond where it was failing this morning (it's running on ursa2 and on my panda board at the minute.)
Ramana
linaro-toolchain@lists.linaro.org