This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.
Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
---|---|---|
Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |
Other format: | [Raw text] |
Hello! We have these long-standing comments in i386.md: "%%% Potential partial reg stall on alternative X. What to do?" Using preferred_for_speed attribute, we can disparage these alternatives for TARGET_PARTIAL_REG_STALL targets, which seems the correct thing to do. 2016-05-07 Uros Bizjak <ubizjak@gmail.com> * config/i386/i386.md (*addqi_1): Add preferred_for_speed attribute to disparage alternatives 3 and 4 for TARGET_PARTIAL_REG_STALL targets. (*andqi_1): Add preferred_for_speed attribute to disparage alternative 2 for TARGET_PARTIAL_REG_STALL targets. (*<code>qi_1): Ditto. (*one_cmplqi2_1): Add preferred_for_speed attribute to disparage alternative 1 for TARGET_PARTIAL_REG_STALL targets. (*ashlqi3_1): Ditto. (*swap<mode>): Merge from *swap<mode>_1 and *swap<mode>_2 patterns. Add preferred_for_size attribute to disparage alternative 0 and preferred_for_speed attribute to disparage alternative 1 for TARGET_PARTIAL_REG_STALL targets. Patch was bootstrapped and regression tested on x86_64-linux-gnu {,-m32}. Committed to mainline SVN. Uros.
Attachment:
p.diff.txt
Description: Text document
Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
---|---|---|
Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |