This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [PATCH]: update to df for mainline.
- From: Bernd Schmidt <bernds_cb1 at t-online dot de>
- To: Kenneth Zadeck <zadeck at naturalbridge dot com>
- Cc: GCC Patches <gcc-patches at gcc dot gnu dot org>, "Berlin, Daniel" <dberlin at dberlin dot org>, Ian Lance Taylor <ian at airs dot com>, Vladimir Makarov <vmakarov at redhat dot com>, "F. Kenneth Zadeck" <kenneth dot zadeck at naturalbridge dot com>
- Date: Tue, 09 May 2006 15:06:09 +0200
- Subject: Re: [PATCH]: update to df for mainline.
- References: <44298EDB.5000804@naturalbridge.com> <4429BE69.5010307@t-online.de> <442AA1D3.4050007@naturalbridge.com>
Dealing with reginfo.diff now...
+ the aggrigate is still needed to service the REG_DEAD and
"aggregate"
- struct df_ri_problem_data *problem_data =
- (struct df_ri_problem_data *) dflow->problem_data;
I think the "=" should be on the second line.
+ struct df * df = dflow->df;
"df *df"
+ return regstack_completed
+ && IN_RANGE (regno, FIRST_STACK_REG, LAST_STACK_REG);
Parenthesize for nicer formatting. Move #ifdefs into the function.
+ if (flags & DF_RI_LIFE)
+ if (df_ignore_stack_reg (REGNO (XEXP (link, 0))))
+ REG_N_DEATHS (REGNO (XEXP (link, 0)))++;
Add break here for clarity.
+ if (all_dead)
+ {
+ struct df_link *regs = mws->regs;
+ rtx note = alloc_EXPR_LIST (REG_UNUSED, *DF_REF_LOC (regs->ref),
+ REG_NOTES (insn));
That probably works, but it looks as if it were using an individual reg
rather than the whole. Maybe this could be written differently.
+ bitmap_set_bit (do_not_gen, regno);
Only the first? Probably ok given what df_set_dead_notes_for_mw does.
Lots of code shared between df_set_{dead,unused}_notes_for_mw.
+ else if ( (!(DF_REF_FLAGS (def) & DF_REF_MW_HARDREG))
Kill the space.
+ rtx note =
+ alloc_EXPR_LIST (REG_UNUSED, reg, REG_NOTES (insn));
Formatting; see above.
+ REG_N_DEATHS (dregno) ++;
Kill the space.
+ /* Kill this register if it is not a subreg store. */
+ if (!(DF_REF_FLAGS (def) & DF_REF_PARTIAL))
+ bitmap_clear_bit (live, dregno);
Looks like this depends on another part. I have to admit that I'm not
convinced yet by the subreg handling in DF as a whole. A normal
(non-multiword non-strict-low) subreg store _does_ kill the whole reg.
+ if ((dflow->flags & DF_RI_LIFE) && (uregno >= FIRST_PSEUDO_REGISTER))
Some parens unnecessary. Elsewhere too.
Do you think we need this patch in stage3, or can it wait until 4.3 stage1?
Bernd