public inbox for gcc-patches@gcc.gnu.org
 help / color / mirror / Atom feed
From: "John David Anglin" <dave@hiauly1.hia.nrc.ca>
To: dave@hiauly1.hia.nrc.ca (John David Anglin)
Cc: gcc-patches@gcc.gnu.org, gcc-bugs@gcc.gnu.org, alan@linuxcare.com.au
Subject: Re: [PATCH] Re: REG_DEAD/REG_EQUIV problem.
Date: Thu, 15 Feb 2001 11:33:00 -0000	[thread overview]
Message-ID: <200102151932.OAA05634@hiauly1.hia.nrc.ca> (raw)
In-Reply-To: <no.id>

> 
> > On Thu, 1 Feb 2001, Alan Modra wrote:
> > 
> > > (insn 17 15 21 (set (reg/v:DI 69)
> > >         (mem/u:DI (lo_sum:DI (reg:DI 70)
> > >                 (unspec:DI[ 
> > >                         (symbol_ref/v/f:DI ("*L$C0001"))
> > >                     ]  0)) 0)) 83 {*pa.md:2368} (insn_list 15 (nil))
> > >     (expr_list:REG_EQUIV (mem/u:DI (lo_sum:DI (reg:DI 70)
> > >                 (unspec:DI[ 
> > >                         (symbol_ref/v/f:DI ("*L$C0001"))
> > >                     ]  0)) 0)
> > >         (expr_list:REG_DEAD (reg:DI 70)
> > >             (nil))))
> > 
> > Whee, maybe I'll be a gcc hacker one day.
> > 
> > This fixes the problem for me.  I've no idea whether it's the ideal
> > solution - it may well be possible and desirable to prevent these
> > conflicting notes being issued.
> > 
> > gcc/ChangeLog
> >	* rtlanal.c (reg_death_note_p): New function.
> > 	(reg_mentioned_dies): New function.
> > 	* rtl.h (reg_mentioned_dies): Declare.
> > 	* reload1.c (reload): Don't set reg_equiv_memory_loc for an insn
> > 	that references memory via a reg that dies.

I am not certain this is totally sufficient.  I think it is possible for this
type of situation to occur:

(set (reg: A) (plus (reg: PIC) (high:SI (symbol_ref:SI ("X")))))

(set (reg: B) (mem/u:SI (lo_sum:SI (reg/f: A) (unspec:SI[(symbol_ref:SI ("X")]))))

(set (reg: C) (mem/u:SI (lo_sum:SI (reg/f: A) (unspec:SI[(symbol_ref:SI ("X")]))))
  (expr_list:REG_DEAD (reg: A) (nil))

...

(set (reg: D) (reg: B))

The last set may be in a different basic block than the one in which B is set.
Reload may try to substitute the MEM for B.  It won't notice that the register
mentioned in the MEM is dead.

It would appear that we need to check whether any registers mentioned in
the MEM are still alive at the point where D is set before substituting.

I have a modified patch which stops the generation of the REG_EQUIV note
in local-alloc (see below).  However, it won't work if the above situation
can occur.

There is also a problem with the above MEM being incorrectly substuted
into the insn at line 2099 of pa.md during reload.  I believe this is because
the Q constraint is relaxed during reload:

/* Subroutine for EXTRA_CONSTRAINT.

   Return 1 iff OP is a pseudo which did not get a hard register and
   we are running the reload pass.  */

#define IS_RELOADING_PSEUDO_P(OP) \
  ((reload_in_progress					\
    && GET_CODE (OP) == REG				\
    && REGNO (OP) >= FIRST_PSEUDO_REGISTER		\
    && reg_renumber [REGNO (OP)] < 0))

/* Optional extra constraints for this machine. Borrowed from sparc.h.

   For the HPPA, `Q' means that this is a memory operand but not a
   symbolic memory operand.  Note that an unassigned pseudo register
   is such a memory operand.  Needed because reload will generate
   these things in insns and then not re-recognize the insns, causing
   constrain_operands to fail.

   `R' is used for scaled indexed addresses.

   `S' is the constant 31.

   `T' is for fp loads and stores.  */
#define EXTRA_CONSTRAINT(OP, C)				\
  ((C) == 'Q' ?						\
   (IS_RELOADING_PSEUDO_P (OP)				\
    || (GET_CODE (OP) == MEM				\
	&& (memory_address_p (GET_MODE (OP), XEXP (OP, 0))\
	    || reload_in_progress)			\
	&& ! symbolic_memory_operand (OP, VOIDmode)	\
        && !(GET_CODE (XEXP (OP, 0)) == PLUS		\
	     && (GET_CODE (XEXP (XEXP (OP, 0), 0)) == MULT\
		 || GET_CODE (XEXP (XEXP (OP, 0), 1)) == MULT))))\
   : ((C) == 'R' ?					\
     (GET_CODE (OP) == MEM				\
      && GET_CODE (XEXP (OP, 0)) == PLUS		\
      && (GET_CODE (XEXP (XEXP (OP, 0), 0)) == MULT	\
	  || GET_CODE (XEXP (XEXP (OP, 0), 1)) == MULT)	\
      && (move_operand (OP, GET_MODE (OP))		\
	  || memory_address_p (GET_MODE (OP), XEXP (OP, 0))\
	  || reload_in_progress))			\


I have modified symbolic_memory_operand to detect pic symbolic references
and would like to get rid of the reload_in_progress stuff in EXTRA_CONSTRAINT.
This seems bogus to me and an attempt to work around some kind of reload
bug.  Maybe Jeff can recall why it is there?

A simple way to avoid the problem is to turn off the unchanging bit in
pic MEMs.  This prevents the REG_EQUIV notes from being generated.  I have
done a complete build with the unchanging bit off.  There are two testsuite
failures that I don't think are present in non PIC builds and checks.  These
are gcc.c-torture/compile/900313-1.c and gcc.c-torture/execute/941014-1.c:

Executing on host: /xxx/gnu/gcc-2.97/objdir/gcc/xgcc -B/xxx/gnu/gcc-2.97/objdir/gcc/ /xxx/gnu/gcc-2.97/gcc/testsuite/gcc.c-torture/compile/900313-1.c  -w  -O0 -c  -fPIC -o /xxx/gnu/gcc-2.97/objdir/gcc/testsuite/900313-1.o    (timeout = -1)
/xxx/gnu/gcc-2.97/gcc/testsuite/gcc.c-torture/compile/900313-1.c: In function `main':
/xxx/gnu/gcc-2.97/gcc/testsuite/gcc.c-torture/compile/900313-1.c:7: Insn does not satisfy its constraints:
(insn 104 55 107 (set (reg:SI 111)
	(reg:SI 1 %r1)) 69 {*pa.md:2099} (nil)
    (nil))
/xxx/gnu/gcc-2.97/gcc/testsuite/gcc.c-torture/compile/900313-1.c:7: confused by earlier errors, bailing out

I will try to figure out why SI 111 didn't get a hard register.

Comments?

Dave
-- 
J. David Anglin                                  dave.anglin@nrc.ca
National Research Council of Canada              (613) 990-0752 (FAX: 952-6605)

2001-02-13  John David Anglin  <dave@hiauly1.hia.nrc.ca>

	* local-alloc.c (update_equiv_regs): Don't change a REG_EQUAL
	note to REG_EQUIV note if the MEM contains a reference to a
	register which dies in the insn.
	* rtlanal.c (reg_death_note_p): New function.
	(reg_mentioned_dies): New function.
	* rtl.h (reg_mentioned_dies): Declare.

--- local-alloc.c.orig	Fri Feb  2 00:54:57 2001
+++ local-alloc.c	Tue Feb 13 17:42:12 2001
@@ -822,6 +822,7 @@
       rtx set;
       rtx dest, src;
       int regno;
+      int mem_is_dead = 0;
 
       if (GET_CODE (insn) == NOTE)
 	{
@@ -956,9 +957,19 @@
       reg_equiv[regno].init_insns
 	= gen_rtx_INSN_LIST (VOIDmode, insn, reg_equiv[regno].init_insns);
 
+      /* Make sure that a MEM doesn't use a register that dies in
+	 this insn.  */
+      if (GET_CODE (src) == MEM)
+	{
+	  rtx death = find_reg_note (insn, REG_DEAD, NULL_RTX);
+	  if (death)
+	    if (reg_mentioned_dies (src, death))
+	      mem_is_dead = 1;
+	}
+
       /* If this register is known to be equal to a constant, record that
 	 it is always equivalent to the constant.  */
-      if (note && ! rtx_varies_p (XEXP (note, 0), 0))
+      if (note && ! mem_is_dead && ! rtx_varies_p (XEXP (note, 0), 0))
 	PUT_MODE (note, (enum machine_mode) REG_EQUIV);
 
       /* If this insn introduces a "constant" register, decrease the priority
--- rtl.h.orig	Sun Jan  7 13:39:19 2001
+++ rtl.h	Mon Feb 12 15:08:06 2001
@@ -1403,6 +1403,7 @@
 extern int computed_jump_p		PARAMS ((rtx));
 typedef int (*rtx_function)             PARAMS ((rtx *, void *));
 extern int for_each_rtx                 PARAMS ((rtx *, rtx_function, void *));
+extern int reg_mentioned_dies		PARAMS ((rtx, rtx));
 extern rtx regno_use_in			PARAMS ((unsigned int, rtx));
 extern int auto_inc_p			PARAMS ((rtx));
 extern void remove_node_from_expr_list	PARAMS ((rtx, rtx *));
--- rtlanal.c.orig	Thu Feb  8 14:05:50 2001
+++ rtlanal.c	Mon Feb 12 15:08:06 2001
@@ -27,6 +27,7 @@
 
 static void set_of_1		PARAMS ((rtx, rtx, void *));
 static void insn_dependent_p_1	PARAMS ((rtx, rtx, void *));
+static int reg_death_note_p	PARAMS ((rtx *, void *));
 
 /* Forward declarations */
 static int computed_jump_p_1	PARAMS ((rtx));
@@ -2327,6 +2328,34 @@
   return 0;
 }
 
+/* Predicate function for reg_mentioned_dies.  */
+
+static int
+reg_death_note_p (reg, notes)
+     rtx *reg;
+     void *notes;
+{
+  rtx link;
+
+  if (GET_CODE (*reg) != REG)
+    return 0;
+
+  for (link = (rtx) notes; link; link = XEXP (link, 1))
+    if (REG_NOTE_KIND (link) == REG_DEAD && *reg == XEXP (link, 0))
+      return 1;
+  return 0;
+}
+
+/* Return 1 if any register found in INSN has a REG_DEAD note in NOTES.  */
+
+int
+reg_mentioned_dies (insn, notes)
+     rtx insn;
+     rtx notes;
+{
+  return for_each_rtx (&insn, reg_death_note_p, notes);
+}
+
 /* Searches X for any reference to REGNO, returning the rtx of the
    reference found if any.  Otherwise, returns NULL_RTX.  */
 

  parent reply	other threads:[~2001-02-15 11:33 UTC|newest]

Thread overview: 506+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <no.id>
1998-12-12 19:15 ` Problem with put_reg_into_stack H.J. Lu
1999-08-06 12:06 ` Internal compiler error in `emit_call_1' - vax-dec-ultrix4.3 John David Anglin
1999-08-25  0:38   ` Jeffrey A Law
1999-08-31 22:41     ` Jeffrey A Law
1999-08-31 22:41   ` John David Anglin
2000-04-28 13:32 ` g++-mike-eh8-C test failure John David Anglin
2000-04-28 15:41   ` John David Anglin
2000-04-30  8:47     ` John David Anglin
2000-05-02  9:30       ` Jeffrey A Law
2000-05-02 11:08         ` John David Anglin
2000-05-02 11:17           ` Jeffrey A Law
2000-05-02 13:02             ` John David Anglin
2000-05-02 18:44               ` Jason Merrill
2000-05-03 10:43                 ` John David Anglin
2000-05-03 15:49                   ` Jason Merrill
2000-05-03 17:27                     ` John David Anglin
2000-05-05 15:12                     ` John David Anglin
2000-05-08 11:10             ` John David Anglin
2000-05-18 11:19 ` VAX Ultrix bootstrap failure: Cannot allocate 4072 bytes John David Anglin
2000-05-18 15:48   ` Jeffrey A Law
2000-05-22  9:45 ` VAX Ultrix bootstrap failure: cc: -o would overwrite John David Anglin
2000-05-22 17:22 ` John David Anglin
2000-05-24 10:32 ` VAX Ultrix bootstrap failure with gcc-2.96 John David Anglin
2000-05-24 11:37   ` Zack Weinberg
2000-05-24 13:08     ` John David Anglin
2000-05-25 18:31       ` Zack Weinberg
2000-05-25 19:21         ` John David Anglin
2000-05-25 19:35           ` Zack Weinberg
2000-05-25 20:05             ` John David Anglin
2000-05-25 20:33               ` Zack Weinberg
2000-06-02 10:31             ` John David Anglin
2000-05-24 14:09     ` John David Anglin
2000-05-24 18:50       ` Jeffrey A Law
     [not found] ` <200005272138.RAA08789@hiauly1.hia.nrc.ca>
2000-05-27 20:19   ` VAX Ultrix bootstrap with gcc-2.96 20000519: genrecog failure Richard Henderson
2000-05-28 11:00     ` John David Anglin
2000-06-02 12:15 ` VAX Ultrix bootstrap failure with gcc-2.96 John David Anglin
2000-06-02 13:09   ` John David Anglin
2000-06-12  8:18 ` Evaluation order of &&s in ||, in macro INDEX_TERM_P, in vax.h John David Anglin
2000-06-13  8:55   ` Jeffrey A Law
2000-06-20  9:55 ` VAX Ultrix bootstrap failure with gcc-2.96 John David Anglin
2000-06-20 11:13   ` Bruce Korb
2000-06-29  9:50 ` collect2: ld terminated with signal 10 [Bus error] John David Anglin
2000-06-30 11:13   ` Jeffrey A Law
2000-08-30 18:03 ` Patches: Re: libio testsuite: timeout compiling tFile.cc John David Anglin
2000-09-01 13:29 ` Patch: Re: objc FAILs under hpux/-threads with gcc-2.96 CVS 20000816 John David Anglin
2000-09-01 22:22   ` Ovidiu Predescu
2000-09-05 11:25     ` Another patch: Re: objc FAILs under hpux/-threads with gcc-2.96 CVS John David Anglin
2000-09-06  0:01       ` Ovidiu Predescu
2000-09-08 10:29 ` Unsatisfied symbols: cpp_register_pragma (code), cpp_register_pragma_space (code) John David Anglin
2000-09-08 10:38   ` Zack Weinberg
2000-09-08 10:43     ` John David Anglin
2000-09-08 11:02       ` Zack Weinberg
2000-09-08 17:15 ` Segmentation fault building libg++ without named returns John David Anglin
2000-09-08 23:06   ` John David Anglin
2000-09-09  6:14     ` Manfred Hollstein
2000-09-09 15:06       ` John David Anglin
2000-09-10  2:55         ` Manfred Hollstein
2000-09-11 13:35           ` John David Anglin
2000-09-11 14:25 ` John David Anglin
2000-09-14  5:45   ` libg++-2.8.1.3-20000914.diff.gz (was: Re: Segmentation fault building libg++ without named returns) Manfred Hollstein
2000-09-21  9:50     ` John David Anglin
2000-09-22 11:06 ` Patch: Include WARN_CFLAGS in CFLAGS passed for building fixinc.sh John David Anglin
2000-09-29 11:12 ` PATCH: Re: libio compilation problem John David Anglin
2000-10-16 13:21 ` PATCH: Upgrade floating comparisons on PA to support unordered operands John David Anglin
2000-11-09  9:40 ` testcase for hppa64 gcc bug John David Anglin
2000-11-09 16:17   ` Alan Modra
2000-12-05 20:11   ` Jeffrey A Law
2000-12-05 20:15     ` John David Anglin
2000-12-05 21:28       ` Alan Modra
2001-01-31 17:17         ` Jeffrey A Law
2000-11-25 17:39 ` PATCH: Re: ../../../libio/stream.cc:60: Internal error: Segmentation fault John David Anglin
2000-11-25 17:40   ` John David Anglin
2000-11-29 21:48     ` Jeffrey A Law
2000-11-30 12:16 ` PATCH: HUGE_VAL should be Infinity John David Anglin
2000-12-03 21:14   ` Jeffrey A Law
2000-12-04 10:52     ` Michael Meissner
2000-12-04 11:25       ` John David Anglin
2000-12-04 11:35         ` Jeffrey A Law
2000-12-06 18:05         ` Jeffrey A Law
2000-12-18 14:18 ` Revised patch: Re: PATCH 1: Re: BOOTSTRAP FAILURE: segementation fault in genattrtab John David Anglin
2000-12-20 19:56 ` V3 PATCH: Some complex<> cleanup (1/2) Robert Lipe
2000-12-20 20:09   ` Benjamin Kosnik
2000-12-20 21:24     ` Robert Lipe
2000-12-20 21:59     ` Robert Lipe
2000-12-20 22:08       ` Benjamin Kosnik
2000-12-20 20:14   ` Gabriel Dos Reis
2000-12-28 22:29 ` problem with target builtin functions Herman ten Brugge
2001-01-14 14:39 ` PATCH: PIC_OFFSET_TABLE_REGNUM_SAVED should be call used when John David Anglin
2001-01-14 17:00   ` Alan Modra
2001-01-14 17:31     ` John David Anglin
2001-01-23  8:32 ` pa.md bugfix John David Anglin
2001-01-24 17:32 ` John David Anglin
2001-01-27 13:46   ` John David Anglin
2001-01-29 21:18     ` Jeffrey A Law
2001-02-15 11:33 ` John David Anglin [this message]
2001-02-16 17:10   ` [PATCH] Re: REG_DEAD/REG_EQUIV problem John David Anglin
2001-03-01 17:31 ` Enum related fixes for gcc build with native cc on vax ultrix John David Anglin
2001-03-01 17:51   ` Richard Henderson
2001-03-02 15:47     ` John David Anglin
2001-03-04 10:28     ` John David Anglin
2001-03-04 11:20       ` Richard Henderson
2001-03-04 10:11 ` cpplib: basename () fix John David Anglin
2001-03-04 17:11 ` Enum fix to cplus-dem.c for gcc build with native cc on vax ultrix John David Anglin
2001-03-06 10:07 ` cpplib: basename () fix John David Anglin
2001-03-06 10:29   ` DJ Delorie
2001-03-06 10:32   ` Zack Weinberg
2001-03-06 10:52     ` John David Anglin
2001-03-06 11:16       ` DJ Delorie
2001-03-06 14:46 ` Patch for wrong number of arguments in call to smallest_mode_for_size John David Anglin
2001-03-16 21:50 ` Patch to rtx_varies_p to improve pic code on PA John David Anglin
2001-04-05 13:41 ` f/ansify.c uses ANSI features John David Anglin
2001-04-12 21:08 ` Sign extension of type with precision of 0 causes fault in force_fit_type John David Anglin
2001-04-14 14:20 ` Where's the axe? Can't walk_tree John David Anglin
2001-04-23 17:02   ` .stabs statements refer to symbol not in source John David Anglin
2001-04-23 17:57     ` John David Anglin
2001-04-21 19:33 ` C++ Issue on GCC 3.0 branch John David Anglin
2001-04-23  2:18   ` Bernd Schmidt
2001-04-23  7:51     ` law
2001-04-23  7:55       ` Bernd Schmidt
2001-04-23  7:56       ` Bernd Schmidt
2001-04-23  8:14         ` law
2001-04-25 10:26   ` Mark Mitchell
2001-04-25 14:04     ` John David Anglin
2001-04-25 17:31       ` Mark Mitchell
2001-04-26  8:32         ` John David Anglin
2001-04-26 10:25           ` Mark Mitchell
2001-04-26 10:02         ` law
2001-05-03  9:57 ` PATCH: Re: jartool.c:539: undefined reference to `strdup' John David Anglin
2001-05-03 10:13   ` Alexandre Oliva
2001-05-03 10:37     ` John David Anglin
2001-05-03 10:53       ` Alexandre Oliva
2001-05-03 10:43     ` Tom Tromey
2001-05-03 11:26     ` John David Anglin
2001-05-03 13:22       ` Tom Tromey
2001-05-03 15:10         ` John David Anglin
2001-05-11 18:32 ` Disappearing labels fix John David Anglin
2001-05-18  8:58   ` John David Anglin
2001-05-18  9:07     ` law
2001-05-14 10:18 ` PATCH: Fix toplev.c breakage on PA after eh merge John David Anglin
2001-05-16 13:27 ` PATCH (revised): " John David Anglin
2001-06-04 10:53 ` PATCH: fix argument promotion John David Anglin
2001-06-09  9:37 ` [v3] build failure from automated checker John David Anglin
2001-06-09 11:44   ` Benjamin Kosnik
2001-06-09 11:55     ` John David Anglin
2001-06-09 12:00       ` Benjamin Kosnik
2001-06-13 16:04         ` John David Anglin
2001-06-13 20:22           ` Alexandre Oliva
2001-06-13 20:49             ` Bruce Korb
2001-06-13 21:41               ` Mark Mitchell
2001-06-13 23:03                 ` Alexandre Oliva
2001-06-13 20:52             ` Bruce Korb
2001-06-13 21:23               ` Alexandre Oliva
2001-06-09 16:57       ` Gabriel Dos Reis
2001-06-09 21:34         ` John David Anglin
2001-06-09 23:22           ` Benjamin Kosnik
2001-06-09 21:21 ` PATCH: gthr-dce.h update for v3 thread compatibility John David Anglin
2001-06-14 18:38 ` PATCH: Fix invalid loader fixups from shared libobjc with John David Anglin
2001-07-10 13:33 ` LO_SUM still breaking rs6000, revert patch? John David Anglin
2001-08-09 14:46 ` ../../gcc/java/class.c:1882: `JCR_SECTION_NAME' undeclared in emit_register_classes John David Anglin
2001-08-09 16:13   ` Richard Henderson
2001-08-09 15:12 ` Simple returns are broken in gcc 3.X John David Anglin
2001-08-09 15:48   ` Richard Henderson
2001-08-22  8:50 ` fix execute/20010518-2.c John David Anglin
2001-08-23 22:55 ` John David Anglin
2001-08-23 22:57   ` Richard Henderson
2001-08-26 14:28     ` as: error 7403: undefined label - _ZTVN10__cxxabiv120__si_class_type_infoE [was Re: fix execute/20010518-2.c] John David Anglin
2001-09-05 22:59 ` CVS Problem: java/parse.c and java/parse-scan.c deleted John David Anglin
2001-09-22 11:35 ` PATCH: pass outgoing float arguments in both floating and general registers in indirect calls using 32 bit ABI John David Anglin
2001-09-24  8:47   ` law
2001-09-24 14:19 ` tiny tree.c update John David Anglin
2001-09-25  5:35   ` Jan Hubicka
2001-09-25  6:54     ` John David Anglin
2001-10-03 12:08 ` PATCH: Check all insns in fallthru to see if label is mentioned John David Anglin
2001-12-04 17:46   ` Richard Henderson
2001-12-08  9:23     ` John David Anglin
2001-12-09 16:12       ` Richard Henderson
2001-10-31 10:39 ` PATCH: Use PLUS instead of HIGH/LO_SUM for large constants - take 2 for PA John David Anglin
2001-11-09 16:21   ` law
2001-11-13  5:27     ` law
2001-11-13 15:03     ` law
2001-11-13 15:03 ` Last alignment change for MEM tracking John David Anglin
2001-11-21 16:04 ` C++ pcc struct return fix John David Anglin
2001-11-30 19:36   ` John David Anglin
2001-12-03  2:52   ` Jason Merrill
2001-12-03  8:40     ` John David Anglin
2001-12-03  8:52       ` Mark Mitchell
2001-12-03 15:19         ` John David Anglin
2001-12-07 11:30           ` Mark Mitchell
2001-12-07 20:55             ` John David Anglin
2001-12-08  3:04               ` Jason Merrill
2001-12-08  9:12                 ` John David Anglin
2001-12-03 10:41       ` Jason Merrill
2001-12-03 11:54         ` John David Anglin
2001-12-01 11:48 ` HPUX 11 "size_t" fixinc problems John David Anglin
2001-12-03 10:37   ` Bruce Korb
2001-12-17 12:39     ` John David Anglin
2001-12-03 14:20 ` Unreviewed C++ patch for PA (HP assembler) John David Anglin
2001-12-03 14:49   ` Benjamin Kosnik
2001-12-04  8:25   ` Jason Merrill
2001-12-04  9:18     ` John David Anglin
2001-12-08  9:29     ` John David Anglin
2001-12-09 19:25       ` Jason Merrill
2001-12-09 19:27       ` Phil Edwards
2001-12-09 14:55 ` PATCH: Check all insns in fallthru to see if label is mentioned John David Anglin
2002-01-08 20:57 ` fix aix -fcprop-register miscompilation John David Anglin
2002-01-10  9:33 ` PATCH: more portable way to fix g77.f-torture/execute/io1.f (was Re: [PATCH] alias.c find_base_value fix) John David Anglin
2002-01-10 16:38   ` John David Anglin
2002-01-14 15:33     ` Toon Moene
2002-01-15 15:35 ` PATCH: Re: ICE in 920624-1.c with -O3 -funroll-loops on vax-dec-ultrix4.3 John David Anglin
2002-01-15 19:31   ` Richard Henderson
2002-01-15 20:27     ` John David Anglin
2002-02-04 16:04     ` John David Anglin
2002-02-04 16:31       ` Richard Henderson
2002-02-04 17:39         ` law
2002-02-04 18:23           ` Richard Henderson
2002-02-04 21:21             ` law
2002-02-04 22:10               ` PATCH: Re: ICE in 920624-1.c with -O3 -funroll-loops on John David Anglin
2002-02-05  0:06                 ` law
2002-02-05  7:34                   ` Jan Hubicka
2002-02-05  8:50                     ` law
2002-02-05 10:40                       ` Jan Hubicka
2002-02-05  6:01                 ` Jan Hubicka
2002-02-05  0:33               ` PATCH: Re: ICE in 920624-1.c with -O3 -funroll-loops on vax-dec-ultrix4.3 Richard Henderson
2002-02-05  0:11                 ` law
2002-02-05  7:08                 ` Jan Hubicka
2002-02-05  7:58                   ` law
2002-02-05  9:32                   ` Richard Henderson
2002-02-05 10:52                     ` Jan Hubicka
2002-02-04 18:01       ` law
2002-01-16  8:53   ` Jan Hubicka
2002-01-16  9:09     ` John David Anglin
2002-01-16  9:58       ` Jan Hubicka
2002-01-21 13:02 ` Fix ld_library_path in g77.exp for hppa64-hp-hpux11.X John David Anglin
2002-01-21 15:22 ` Fix predicate in decrement_and_branch_until_zero pattern on PA John David Anglin
2002-01-21 15:24   ` Richard Henderson
2002-01-21 15:29     ` law
2002-01-21 16:24       ` Richard Henderson
2002-01-21 15:32     ` John David Anglin
2002-01-21 16:27       ` Richard Henderson
2002-01-21 21:58         ` John David Anglin
2002-01-21 22:21           ` Richard Henderson
2002-01-21 22:48         ` John David Anglin
2002-01-22 23:30 ` PATCH: Fix loop.c for targets without HAVE_prefetch John David Anglin
2002-01-23 12:15   ` H . J . Lu
2002-01-28  9:28 ` gcc failed to bootstrap on Linux/mipsel John David Anglin
2002-01-28 11:10   ` Richard Henderson
2002-01-28 11:18     ` John David Anglin
2002-01-28 11:19       ` Richard Henderson
2002-01-28 11:39         ` John David Anglin
2002-01-28 12:07           ` Richard Henderson
2002-01-28 16:04             ` John David Anglin
2002-01-28 17:00               ` Richard Henderson
2002-02-04 13:02 ` Define _GNU_SOURCE in unwind-dw2-fde-glibc.c John David Anglin
2002-02-04 13:43   ` Richard Henderson
2002-02-05 21:17 ` Add missing predicate to PREDICATE_CODES on PA John David Anglin
2002-02-05 23:20   ` Graham Stott
2002-02-06  9:06     ` John David Anglin
2002-02-10 12:30     ` John David Anglin
2002-02-16 16:12 ` Don't use lib2funcs.asm " John David Anglin
2002-02-16 17:56   ` law
2002-02-16 19:45     ` John David Anglin
2002-02-17  9:02       ` law
2002-02-17 11:29         ` John David Anglin
2002-02-18  3:21           ` Olivier Hainque
2002-02-18  9:26             ` John David Anglin
2002-02-18 10:23               ` Olivier Hainque
2002-02-18 10:27                 ` John David Anglin
2002-02-19  2:35                   ` Olivier Hainque
2002-02-18 12:18           ` Hans-Peter Nilsson
2002-02-18 12:30 ` John David Anglin
2002-02-19  2:24   ` Olivier Hainque
2002-03-13  8:34 ` fix for strct-pack-1.c regressions John David Anglin
2002-03-28 15:44 ` Letext John David Anglin
2002-05-08 13:54 ` [PATCH] checking version of rtl flag access macros John David Anglin
2002-06-03 14:02 ` [PATCH] Jump bypassing and improved cprop (take 2) John David Anglin
2002-06-04  6:31   ` law
2002-06-04  6:45     ` Jan Hubicka
2002-06-04  9:09       ` John David Anglin
2002-06-04  9:12         ` Jan Hubicka
2002-06-05 21:30           ` John David Anglin
2002-06-05 23:19             ` law
2002-06-05 14:23 ` Patch: Use tm_defines to configure default scheduling model on PA John David Anglin
2002-06-12 10:03 ` John David Anglin
2002-06-12 10:23   ` DJ Delorie
2002-06-12 10:51     ` John David Anglin
2002-06-12 11:17       ` DJ Delorie
2002-06-12 11:47         ` John David Anglin
2002-06-12 12:01           ` DJ Delorie
2002-06-12 13:01             ` John David Anglin
2002-06-15 11:04     ` John David Anglin
2002-06-18 16:01 ` [PATCH/RFA] Allow register other than SP for DWARF2 CFA John David Anglin
2002-06-24 12:48 ` PATH: inline does not work with -O3 specified Herman ten Brugge
2002-06-30 16:59   ` Michael Hayes
2002-07-11 12:17 ` PATCH: Fix failure of arith-rand-ll.c on hppa 32-bit targets John David Anglin
2002-08-02 22:10 ` [RFA] Fix libsupc++/Makefile.in John David Anglin
2002-08-02 23:49   ` Neil Booth
2002-08-21  9:31 ` PATCH: fix warning and return value for remove_dup_nonsys_dirs John David Anglin
2002-08-21 10:13   ` Zack Weinberg
2002-08-31  9:24 ` 128 bit floats on PA64 John David Anglin
2002-09-23 11:53 ` Patch for PR c/4319 John David Anglin
2002-09-30 21:03 ` PATCH for sibcalls on i386 John David Anglin
2002-10-21 15:48 ` [PATCH] Inline __udiv_w_sdiv into __divdi3 etc John David Anglin
2002-10-24 12:51 ` Reload patch for PA call rewrite John David Anglin
2002-10-24 16:26 ` John David Anglin
2002-10-25  8:54   ` Jeff Law
2002-10-25 10:10     ` John David Anglin
2002-10-30 19:23 ` Call rewrite for PA John David Anglin
2002-11-09 15:18 ` C++ PATCH: ABI bug for vcall offsets John David Anglin
2002-11-09 15:27   ` John David Anglin
2002-11-10 18:21   ` Mark Mitchell
2002-11-10 19:42     ` John David Anglin
2002-11-13 13:12 ` gcc-64 20021111 broken on HP-UX 11.00 John David Anglin
2002-11-26 11:28 ` Unreviewed patch John David Anglin
2002-11-26 15:53 ` PATCH: Fix handling of return values handled in PARALLELs John David Anglin
2002-11-26 17:58   ` Richard Henderson
2002-11-26 19:52 ` clean up some hook routines John David Anglin
2002-12-04 14:22 ` HP-UX PA long double alignment change John David Anglin
2002-12-04 14:31   ` Steve Ellcey
2002-12-04 18:14 ` John David Anglin
2003-01-17 17:15 ` [PATCH] Fix RTL sharing problem in CSE John David Anglin
2003-01-17 17:24   ` law
2003-01-17 18:25     ` Jan Hubicka
2003-01-17 18:59       ` Roger Sayle
2003-01-17 22:33         ` David Edelsohn
2003-01-17 23:56           ` Dale Johannesen
2003-01-17 22:46         ` law
2003-01-19 16:59 ` [PATCH]: Fix ICE in convert_move on PA John David Anglin
2003-01-20 17:31   ` Richard Henderson
2003-01-24  3:44 ` [PATCH] Fix find_reloads_address bug, take 2 Ulrich Weigand
2003-02-01 21:40 ` ping: Unreviewed patch to fix patch to fix bootstrap failure on PA John David Anglin
2003-02-01 21:43   ` Zack Weinberg
2003-02-02  0:23   ` Geoff Keating
2003-02-03  5:02 ` hppa-linux regressions and 3.2.2 release John David Anglin
2003-02-03 11:03   ` Gabriel Dos Reis
2003-02-03 16:26   ` John David Anglin
2003-02-03 16:54     ` Gabriel Dos Reis
2003-02-03 18:02       ` John David Anglin
2003-02-04 21:20 ` Mainline bootstrap failure on hppa2.0w-hp-hpux11.00 John David Anglin
2003-02-05 18:21 ` John David Anglin
2003-02-05 18:46 ` John David Anglin
2003-02-05 19:12   ` Kaveh R. Ghazi
2003-02-11 19:37 ` Bootstrap failure on hppa-unknown-linux-gnu, trunk John David Anglin
2003-02-11 22:37   ` Josef Zlomek
2003-02-11 22:51     ` John David Anglin
2003-03-05 22:00   ` Josef Zlomek
2003-03-05 22:03     ` Josef Zlomek
2003-03-11  2:04 ` jcf-io.c:339: warning: `origsep' might be used uninitialized John David Anglin
2003-03-21  0:02 ` [PATCH]: PA long unconditional branch generation (PR10062) John David Anglin
2003-04-10 19:52 ` Failure of test07 in 27_io/filebuf_members.cc under HP-UX (PR 9964) John David Anglin
2003-04-11  1:22   ` Benjamin Kosnik
2003-04-18 16:32 ` PATCH: Fix PR 8866 John David Anglin
2003-05-10  2:15 ` [PATCH] allow zero_extract combines - checked in John David Anglin
2003-05-12  2:26   ` Eric Christopher
2003-05-12  2:43     ` John David Anglin
2003-05-12  9:11       ` Eric Christopher
2003-05-12 16:50   ` Eric Christopher
2003-05-12 17:05     ` John David Anglin
2003-05-12 17:13       ` Eric Christopher
2003-05-12 17:59         ` John David Anglin
2003-05-12 18:03           ` Eric Christopher
2003-05-23  1:17 ` speedup collect2 (by not using it) John David Anglin
2003-06-19 22:29 ` [PATCH] Fix PCH failures on SPARC John David Anglin
2003-06-20 19:12 ` Hookize CFG debugging code John David Anglin
2003-06-20 19:35   ` Jan Hubicka
2003-06-20 20:47 ` [PATCH]: Fix label replacement in REG_NOTES John David Anglin
2003-07-03 20:17 ` Bootstrap failure compiling ada/misc.c John David Anglin
2003-07-04 17:54 ` [Boehm-GC] Limit +ESdbgasm to HPUX cc on PA John David Anglin
2003-07-04 18:16   ` Tom Tromey
2003-07-04 18:37     ` John David Anglin
2003-07-04 20:24 ` Unreviewed fix for bootstrap failure John David Anglin
2003-10-06 16:53 ` [PATCH] Bootstrap failure due to reload bug Ulrich Weigand
2003-11-01 23:48 ` [PATCH] Fix AMD64 handling of functions with huge stack frames (take 2) John David Anglin
2004-01-22 21:43 ` [PATCH] Optimize subregs of zero and sign extensions " John David Anglin
2004-02-28  5:22 ` [patch] do not disregard LD_LIBRARY_PATH for c++, g77 and objc tests John David Anglin
2004-02-28 18:12   ` Geoff Keating
2004-02-28 18:30     ` Eric Botcazou
2004-04-16 22:12 ` [committed 3.5] Tweak xfail for gcc.dg/const-elim-1.c John David Anglin
2004-04-17 19:09   ` Mark Mitchell
2004-04-18 22:01 ` [patch 3.3/3.4/3.5] Fix PR bootstrap/14671 John David Anglin
2004-04-18 22:51   ` Mark Mitchell
2004-04-18 23:04     ` John David Anglin
2004-04-19  3:24 ` John David Anglin
2004-04-19  3:27   ` Mark Mitchell
2004-04-19  5:30   ` Zdenek Dvorak
2004-04-19 13:58     ` John David Anglin
2004-04-19 14:49     ` Mark Mitchell
2004-04-19 15:01       ` Zdenek Dvorak
2004-04-19 16:22         ` John David Anglin
2004-04-19 17:45           ` Zdenek Dvorak
2004-04-19 19:57         ` John David Anglin
2004-04-19 20:03           ` Andrew Pinski
2004-04-20 16:05             ` John David Anglin
2004-04-21 19:07               ` Mark Mitchell
2004-04-21 21:44               ` Richard Henderson
2004-04-22 20:38                 ` Gabriel Dos Reis
2004-04-21 19:23 ` [committed 3.5] Fix DBX register numbering for hppa64 John David Anglin
2004-04-22 17:25   ` Mark Mitchell
2004-04-22 17:40     ` John David Anglin
2004-04-25 16:00 ` [PATCH] Re: Fix problem with constant modulus John David Anglin
2004-04-25 22:17 ` [ping] Unreviewed patch John David Anglin
2004-04-27 23:34   ` Mark Mitchell
2004-05-04  0:50 ` [PATCH] Ulrich Weigand
2004-05-04  0:52   ` [PATCH] Eric Christopher
2004-06-04 15:56 ` building sh-elf / sh-linux (Was: Re: [PATCH/RFA] PR target/13250) Joern Rennecke
2004-06-04 18:25   ` Joern Rennecke
2004-07-07 21:10 ` [PATCH] Fix PR target/16344 John David Anglin
2004-07-08  5:56   ` Mark Mitchell
2004-07-09 21:01 ` [RFT/RFA] gimplify pa va_arg John David Anglin
2004-07-10 18:38 ` gimple va_arg for hppa John David Anglin
2004-07-11 13:07 ` [PATCH] DWARF-2 unwinder off-by-one problem with signal frames Ulrich Weigand
2004-10-31 20:11 ` [PATCH] Fix PR target/16304: AIX 4.x forward reference proble John David Anglin
2004-11-27 17:39 ` [committed] Fix pch/14940 on hppa-unknown-linux-gnu John David Anglin
2004-11-28 20:24 ` [ping] Fix PR target/16304: AIX 4.x forward reference problem John David Anglin
2004-12-05  0:01 ` [patch] Fix for PR 14838 John David Anglin
2004-12-05  1:17   ` Richard Henderson
2004-12-05  5:18   ` Gabriel Dos Reis
2004-12-27  2:57 ` [committed] Fix PR target/17643 on main and 3.4, and 3.3 John David Anglin
2005-01-21  0:25 ` Change to gcc.dg/tree-ssa/loop-1.c John David Anglin
2005-01-21  0:31   ` John David Anglin
2005-01-21  0:43     ` John David Anglin
2005-01-21  0:49       ` Steve Ellcey
2005-01-21 14:57         ` John David Anglin
2005-01-21 16:44           ` Steve Ellcey
2005-01-21 17:42             ` John David Anglin
2005-01-21 20:33               ` Janis Johnson
2005-04-06 16:49 ` [patch] Add Ada tasking support for hppa-unknown-linux-gnu (take 2) John David Anglin
2005-04-29 10:27   ` Arnaud Charlet
2005-04-29 13:43     ` John David Anglin
2005-04-29 15:07       ` Arnaud Charlet
2005-04-29 15:19         ` John David Anglin
2005-04-29 23:41         ` Richard Henderson
2005-05-01 22:52           ` Mark Mitchell
2005-05-02 17:49             ` Florian Weimer
2005-05-02 18:54               ` Mark Mitchell
2005-04-30  0:14     ` John David Anglin
2005-05-02  9:54       ` Arnaud Charlet
2005-05-02 13:40         ` John David Anglin
2005-05-02 13:44           ` Arnaud Charlet
2005-05-02 15:45             ` John David Anglin
2005-05-02 15:48               ` Arnaud Charlet
2005-05-02 16:05                 ` John David Anglin
2005-05-03  8:22                   ` Arnaud Charlet
2005-06-13  1:33         ` John David Anglin
2005-06-13  6:31           ` Arnaud Charlet
2005-06-13 14:16             ` John David Anglin
2005-06-13 15:22               ` Arnaud Charlet
2005-08-20 16:15 ` [patch]: Fix PR testsuite/23239 John David Anglin
2005-12-07  0:52 ` fix post-weakref gthr-*.h on HP-UX (and others?) John David Anglin
2006-01-17  4:54 ` [PING * 2] One line patch John David Anglin
2006-01-17  4:58 ` [committed] Fix PR target/20754: ACATS cxg1005 fails at runtime on hppa-linux John David Anglin
2006-02-05 16:29 ` [committed] Fix PR target/25926: A87B59A SIGABRT John David Anglin
2006-11-18 18:11 ` Ping: [PATCH] Limit precision of *bitsizetypes types John David Anglin
2006-11-18 21:47   ` Roger Sayle
2006-11-18 21:51     ` John David Anglin
2006-11-18 23:27       ` Roger Sayle
2007-01-25 23:55 ` [PATCH, commited] PR other/30182, Fix __builtin_finite on HP-UX John David Anglin
2007-01-26  0:58   ` Steve Ellcey
2007-02-28 10:53 ` [committed] Fix bug target/30634 John David Anglin
2007-03-17  0:31 ` [committed] Fix long local calls on PA HP-UX SOM target John David Anglin
2007-08-27  2:07 ` [patch,testsuite] Fix PR testsuite/33153 John David Anglin
2007-08-27  8:56   ` Jakub Jelinek
2007-08-28  5:31     ` John David Anglin
2007-09-13  2:28     ` John David Anglin
2007-09-13  5:45       ` Mark Mitchell
2007-09-16  9:20 ` ping: [PATCH] Fix PR middle-end/33273 John David Anglin
2007-09-16  9:21   ` Jakub Jelinek
2007-09-16 19:49   ` Richard Guenther
2007-12-10  3:24 ` [committed] [PR target/34091] Secondary reloads for floating-point register classes John David Anglin
2007-12-14  1:57 ` John David Anglin
2007-12-22 11:05 ` [committed] Fix PR target/34525 - unrecognized insn (take 2) John David Anglin
2008-01-18 10:22 ` [ping] Ignore thread local symbols when generating dbx debug info John David Anglin
2008-02-08  2:49   ` John David Anglin
2008-03-05  3:28     ` Jim Wilson
2008-03-08 20:48       ` [ping] Ignore thread local symbols when generating dbx debug John David Anglin
2008-03-10 16:27         ` [Bulk] " Jim Wilson
2008-03-10 16:45           ` [Bulk] Re: [ping] Ignore thread local symbols when generating John David Anglin
2008-03-12  6:32             ` Jim Wilson
2008-02-07  2:32 ` [committed] Remove xfail for hppa*-*-* from g++.dg/tree-ssa/ivopts-1.C John David Anglin
2008-02-09 22:03 ` RFC: Fix PR middle-end/34150 -- Lost LABEL_NUSES counts John David Anglin
2008-08-18 14:23 ` update dwarf2 asm unwind info [hppa64-*-* failures] John David Anglin
2008-08-21 20:10   ` Richard Henderson
2008-08-21 21:28     ` John David Anglin
2008-08-22  0:09       ` Richard Henderson
2008-08-22  3:27         ` John David Anglin
2008-08-22 17:53           ` Richard Henderson
2008-08-22 18:30             ` John David Anglin
2008-08-22 20:57               ` Richard Henderson
2008-08-23 19:46       ` John David Anglin
2008-08-23 21:14         ` John David Anglin
2008-08-24 20:09           ` Richard Henderson
2008-08-24 21:06             ` John David Anglin
2008-08-25  1:42               ` Richard Henderson
2008-08-25  3:31                 ` John David Anglin
2008-08-25  4:57                 ` John David Anglin
2009-08-02 19:35 ` [committed] Fix previous change to pa.c John David Anglin
2009-11-25  3:28 ` [committed] Shorten non PIC PA 1.1 calls on hppa-hpux John David Anglin
2010-03-27 15:44 ` [PATCH] Fix visibility of constructors/destructors with -fwhole-program John David Anglin
2010-03-27 15:52   ` Richard Guenther
2010-03-27 17:53   ` Jan Hubicka
2010-03-27 21:11     ` [PATCH] Fix visibility of constructors/destructors with John David Anglin
2010-03-27 21:48       ` Jan Hubicka
2010-03-27 21:54         ` Richard Guenther
2010-03-28  1:39           ` Jan Hubicka
2001-02-02 20:27 [PATCH] Re: REG_DEAD/REG_EQUIV problem John David Anglin
     [not found] <Pine.LNX.4.21.0102011651560.19209-100000@front.linuxcare.com.au>
2001-02-01  1:47 ` Alan Modra
2001-02-17 23:12   ` Jeffrey A Law
     [not found] <no.id@sources.redhat.com>

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=200102151932.OAA05634@hiauly1.hia.nrc.ca \
    --to=dave@hiauly1.hia.nrc.ca \
    --cc=alan@linuxcare.com.au \
    --cc=gcc-bugs@gcc.gnu.org \
    --cc=gcc-patches@gcc.gnu.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).