| Message ID | 5277F436.6040601@redhat.com |
|---|---|
| State | New |
| Headers | show |
On Mon, Nov 4, 2013 at 2:23 PM, Vladimir Makarov <vmakarov@redhat.com> wrote: > The following patch fixes > > http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58967 > > The removed code is too old. To be honest, I even don't remember why I > added this. LRA has been changed a lot since this change and now it > works fine without it. > > There is no test for this case as it is too big. > > The patch was successfully bootstrapped on ppc64 (with LRA) and tested > on ppc{32|64}. > > Is it ok to commit it to the trunk? > > 2013-11-04 Vladimir Makarov <vmakarov@redhat.com> > > PR rtl-optimization/58967 > * config/rs6000/rs6000.c (legitimate_lo_sum_address_p): Remove > !lra_in_progress for mode sizes bigger word. > > > Index: config/rs6000/rs6000.c > =================================================================== > --- config/rs6000/rs6000.c (revision 204305) > +++ config/rs6000/rs6000.c (working copy) > @@ -6388,7 +6388,7 @@ legitimate_lo_sum_address_p (enum machin > return false; > if (GET_MODE_NUNITS (mode) != 1) > return false; > - if (! lra_in_progress && GET_MODE_SIZE (mode) > UNITS_PER_WORD > + if (GET_MODE_SIZE > (mottp://gcc.gnu.org/bugzilla/show_bug.cgi?id=58967de) > UNITS_PER_WORD This is okay, assuming the weird bugzilla URL paste shown in the patch is not really included. Thanks, David
On 11/4/2013 2:23 PM, Vladimir Makarov wrote: > The following patch fixes > > http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58967 > > The removed code is too old. To be honest, I even don't remember why I > added this. LRA has been changed a lot since this change and now it > works fine without it. Whenever I see a comment like this, it reminds me to remind everyone to comment your code! Do not assume you will remember why you wrote what you wrote, so even if it is you who will look at your code, write comments for yourself assuming you have totally forgotten!
On 11/04/2013 02:26 PM, David Edelsohn wrote: > On Mon, Nov 4, 2013 at 2:23 PM, Vladimir Makarov <vmakarov@redhat.com> wrote: >> The following patch fixes >> >> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58967 >> >> The removed code is too old. To be honest, I even don't remember why I >> added this. LRA has been changed a lot since this change and now it >> works fine without it. >> >> There is no test for this case as it is too big. >> >> The patch was successfully bootstrapped on ppc64 (with LRA) and tested >> on ppc{32|64}. >> >> Is it ok to commit it to the trunk? >> >> 2013-11-04 Vladimir Makarov <vmakarov@redhat.com> >> >> PR rtl-optimization/58967 >> * config/rs6000/rs6000.c (legitimate_lo_sum_address_p): Remove >> !lra_in_progress for mode sizes bigger word. >> >> >> Index: config/rs6000/rs6000.c >> =================================================================== >> --- config/rs6000/rs6000.c (revision 204305) >> +++ config/rs6000/rs6000.c (working copy) >> @@ -6388,7 +6388,7 @@ legitimate_lo_sum_address_p (enum machin >> return false; >> if (GET_MODE_NUNITS (mode) != 1) >> return false; >> - if (! lra_in_progress && GET_MODE_SIZE (mode) > UNITS_PER_WORD >> + if (GET_MODE_SIZE >> (mottp://gcc.gnu.org/bugzilla/show_bug.cgi?id=58967de) > UNITS_PER_WORD > This is okay, assuming the weird bugzilla URL paste shown in the patch > is not really included. > > Sorry, it was some garbage occurred when I composed the email. Committed as rev. 204365.
Index: config/rs6000/rs6000.c =================================================================== --- config/rs6000/rs6000.c (revision 204305) +++ config/rs6000/rs6000.c (working copy) @@ -6388,7 +6388,7 @@ legitimate_lo_sum_address_p (enum machin return false; if (GET_MODE_NUNITS (mode) != 1) return false; - if (! lra_in_progress && GET_MODE_SIZE (mode) > UNITS_PER_WORD + if (GET_MODE_SIZE (mottp://gcc.gnu.org/bugzilla/show_bug.cgi?id=58967de) > UNITS_PER_WORD && !(/* ??? Assume floating point reg based on mode? */ TARGET_HARD_FLOAT && TARGET_FPRS && TARGET_DOUBLE_FLOAT
The following patch fixes http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58967 The removed code is too old. To be honest, I even don't remember why I added this. LRA has been changed a lot since this change and now it works fine without it. There is no test for this case as it is too big. The patch was successfully bootstrapped on ppc64 (with LRA) and tested on ppc{32|64}. Is it ok to commit it to the trunk? 2013-11-04 Vladimir Makarov <vmakarov@redhat.com> PR rtl-optimization/58967 * config/rs6000/rs6000.c (legitimate_lo_sum_address_p): Remove !lra_in_progress for mode sizes bigger word. && (mode == DFmode || mode == DDmode)))