simple pentium4 support

Joseph S. Myers jsm28@cam.ac.uk
Tue Feb 27 15:45:00 GMT 2001


On Tue, 27 Feb 2001, Jan Hubicka wrote:

> Tue Feb 27 14:22:01 CET 2001  Jan Hubicka  <jh@suse.cz>
> 	* i386.c (pentium4_cost): New.
> 	(m_PENT4): New macro.
> 	(x86_push_memory, x86_movx,x86_cmove, x86_deep_branch, x86_use_sahf
> 	x86_sub_esp_4, x86_sub_esp_8, x86_add_esp_4, x86_add_esp_8
> 	x86_integer_DFmode_moves, x86_partial_reg_dependency,
> 	x86_memory_mismatch_stall): Add Pentium4
> 	(x86_use_q_reg, x86_use_any_reg): Kill.
> 	(override_options): Add pentium4.
> 	(incdec_operand): Return 0 for pentium4.
> 	(ix86_issue_rate): Add PROCESSOR_PENTIUM4 and PROCESSOR_ATHLON.
> 	* i386.h (x86_use_q_reg, x86_use_any_reg): Kill.
> 	(TARGET_PENTIUM4): Define.
> 	(enum processor_type): Add PROCESSOR_PENTIUM4.
> 	(CPP_CPU_DEFAULT_SPEC): Add pentium4 support.
> 	* i386.md (attribute "cpu"): Add pentium4.

invoke.texi needs to be updated as well.

I guess an update to config.gcc to allow TARGET_CPU_DEFAULT to be set for
Pentium 4 would be harmless as well - but would require appropriate cases
in config.sub to understand configuring for pentium4-*.  (At present, it
also seems that config.sub would handle configuring for athlon-*, but
TARGET_CPU_DEFAULT wouldn't get set in that case - so config.gcc should
perhaps be updated for that as well.)

-- 
Joseph S. Myers
jsm28@cam.ac.uk



More information about the Gcc-patches mailing list