This is an automated email from the git hooks/post-receive script. It was generated because a ref change was pushed to the repository containing the project "CMU Common Lisp". The branch, master has been updated via 86f6dccd0b106765670265e0cf0c7ce2e6a84b55 (commit) from 0d6f20340333a05b1f990056b38ffb77083de2f1 (commit) Those revisions listed above that are new to this repository have not appeared on any other notification email; so we list those revisions in full, below. - Log ----------------------------------------------------------------- commit 86f6dccd0b106765670265e0cf0c7ce2e6a84b55 Author: Raymond Toy <toy.raymond@gmail.com> Date: Sat Jun 2 15:02:00 2012 -0700 Add comments about using clang instead of gcc on x86. diff --git a/src/lisp/Config.x86_common b/src/lisp/Config.x86_common index 5ed23de..0fa9bb3 100644 --- a/src/lisp/Config.x86_common +++ b/src/lisp/Config.x86_common @@ -45,7 +45,24 @@ ifdef FEATURE_UNICODE CPP_DEFINE_OPTIONS += -DUNICODE endif +# Default to using gcc CC = gcc + +# But we can use clang. +# +# However, clang seems to want to use SSE instructions in various +# places, but we DON'T want that because we need a lisp that will run +# on chips without sse. +# +# But on Mac, every machine has SSE2 so we can use SSE2. However, +# there's some code path through GC or allocation where we aren't +# saving the FPU state so after GC or allocation, some XMM FP +# registers are corrupted. +# +# Got that? + +#CC = clang -mno-sse + LD = ld ifeq ($(filter 2% 3%, $(shell $(CC) -dumpversion)),) @@ -70,3 +87,7 @@ DEPEND_FLAGS = # -ffloat-store and -fno-strict-aliasing anymore. e_rem_pio2.o : e_rem_pio2.c $(CC) -c $(CFLAGS) $(CPPFLAGS) $< + +# +gencgc.o : gencgc.c + $(CC) -c -mno-sse $(CFLAGS) $(CPPFLAGS) $< \ No newline at end of file ----------------------------------------------------------------------- Summary of changes: src/lisp/Config.x86_common | 21 +++++++++++++++++++++ 1 files changed, 21 insertions(+), 0 deletions(-) hooks/post-receive -- CMU Common Lisp