public inbox for gcc-patches@gcc.gnu.org
 help / color / mirror / Atom feed
From: Jakub Jelinek <jakub@redhat.com>
To: Jason Merrill <jason@redhat.com>
Cc: "Joseph S. Myers" <joseph@codesourcery.com>,
	       Richard Henderson <rth@redhat.com>,
	Tom Tromey <tromey@redhat.com>,
	       Jan Kratochvil <jkratoch@redhat.com>,
	gcc-patches@gcc.gnu.org,
	       Roland McGrath <roland@redhat.com>
Subject: Re: [RFC PATCH] -grecord-gcc-switches (PR other/32998)
Date: Thu, 14 Jul 2011 09:48:00 -0000	[thread overview]
Message-ID: <20110714094148.GB2687@tyan-ft48-01.lab.bos.redhat.com> (raw)
In-Reply-To: <4E1DAC22.1050709@redhat.com>

On Wed, Jul 13, 2011 at 10:30:58AM -0400, Jason Merrill wrote:
> On 07/13/2011 10:06 AM, Jakub Jelinek wrote:
> >>>--- gcc/testsuite/lib/dg-pch.exp.jj	2011-01-03 18:58:03.000000000 +0100
> >>>+++ gcc/testsuite/lib/dg-pch.exp	2011-07-12 23:13:50.943670171 +0200
> >>>-	dg-test -keep-output "./$bname$suffix" "$otherflags $flags" ""
> >>>+	dg-test -keep-output "./$bname$suffix" "-gno-record-gcc-switches $otherflags $flags" ""
> >>
> >It is only necessary if somebody wants to make -grecord-gcc-switches
> >the default (for bootstrap/regtest I've tweaked common.opt to do that
> >to test it better).  PCH is a big mess and screws debuginfo in many ways,
> >in this case it was just small differences in DW_AT_producer, but
> >we have e.g. ICEs with PCH and -feliminate-dwarf-dups etc.
> 
> Why would PCH change DW_AT_producer?  Because we're restoring
> single_comp_unit_die from the PCH?  Then perhaps we should set
> DW_AT_producer in output_comp_unit rather than gen_compile_unit_die.

output_comp_unit is too late, because at that point has the form been
finalized and sizes calculated etc.  But at the start of dwarf2out_finish
it is possible.
Here is a PCH friendly variant of the patch which tries to set right
producer from the start, but early in dwarf2out_finish double checks
if PCH hasn't changed it and if it did, updates it back to the expected
string.

2011-07-14  Jakub Jelinek  <jakub@redhat.com>

	PR other/32998
	* common.opt (grecord-gcc-switches, gno-record-gcc-switches): New
	options.
	* dwarf2out.c: Include opts.h.
	(dchar_p): New typedef.  Define heap VEC for it.
	(producer_string): New variable.
	(gen_producer_string): New function.
	(gen_compile_unit_die): Use it.
	(dwarf2out_finish): Fix up comp_unit_die () DW_AT_producer
	if needed.
	* Makefile.in (dwarf2out.o): Depend on $(OPTS_H).

--- gcc/common.opt.jj	2011-07-13 17:31:09.000000000 +0200
+++ gcc/common.opt	2011-07-14 10:36:40.000000000 +0200
@@ -2184,6 +2184,14 @@ ggdb
 Common JoinedOrMissing
 Generate debug information in default extended format
 
+gno-record-gcc-switches
+Common RejectNegative Var(dwarf_record_gcc_switches,0) Init(0)
+Don't record gcc command line switches in DWARF DW_AT_producer.
+
+grecord-gcc-switches
+Common RejectNegative Var(dwarf_record_gcc_switches,1)
+Record gcc command line switches in DWARF DW_AT_producer.
+
 gstabs
 Common JoinedOrMissing Negative(gstabs+)
 Generate debug information in STABS format
--- gcc/dwarf2out.c.jj	2011-07-13 17:31:18.000000000 +0200
+++ gcc/dwarf2out.c	2011-07-14 11:04:50.000000000 +0200
@@ -94,6 +94,7 @@ along with GCC; see the file COPYING3.  
 #include "tree-pass.h"
 #include "tree-flow.h"
 #include "cfglayout.h"
+#include "opts.h"
 
 static void dwarf2out_source_line (unsigned int, const char *, int, bool);
 static rtx last_var_location_insn;
@@ -18108,13 +18109,125 @@ gen_ptr_to_mbr_type_die (tree type, dw_d
   add_type_attribute (ptr_die, TREE_TYPE (type), 0, 0, context_die);
 }
 
+typedef const char *dchar_p; /* For DEF_VEC_P.  */
+DEF_VEC_P(dchar_p);
+DEF_VEC_ALLOC_P(dchar_p,heap);
+
+static char *producer_string;
+
+/* Return a heap allocated producer string including command line options
+   if -grecord-gcc-switches.  */
+
+static char *
+gen_producer_string (void)
+{
+  size_t j;
+  VEC(dchar_p, heap) *switches = NULL;
+  const char *language_string = lang_hooks.name;
+  char *producer, *tail;
+  const char *p;
+  size_t len = dwarf_record_gcc_switches ? 0 : 3;
+  size_t plen = strlen (language_string) + 1 + strlen (version_string);
+
+  for (j = 1; dwarf_record_gcc_switches && j < save_decoded_options_count; j++)
+    switch (save_decoded_options[j].opt_index)
+      {
+      case OPT_o:
+      case OPT_d:
+      case OPT_dumpbase:
+      case OPT_dumpdir:
+      case OPT_auxbase:
+      case OPT_auxbase_strip:
+      case OPT_quiet:
+      case OPT_version:
+      case OPT_v:
+      case OPT_w:
+      case OPT_L:
+      case OPT_D:
+      case OPT_I:
+      case OPT_U:
+      case OPT_SPECIAL_unknown:
+      case OPT_SPECIAL_ignore:
+      case OPT_SPECIAL_program_name:
+      case OPT_SPECIAL_input_file:
+      case OPT_grecord_gcc_switches:
+      case OPT_gno_record_gcc_switches:
+      case OPT__output_pch_:
+      case OPT_fdiagnostics_show_location_:
+      case OPT_fdiagnostics_show_option:
+      case OPT____:
+      case OPT__sysroot_:
+	/* Ignore these.  */
+	continue;
+      default:
+	if (save_decoded_options[j].orig_option_with_args_text[0] != '-')
+	  continue;
+	switch (save_decoded_options[j].orig_option_with_args_text[1])
+	  {
+	  case 'M':
+	  case 'i':
+	  case 'W':
+	    continue;
+	  case 'n':
+	    if (save_decoded_options[j].orig_option_with_args_text[2] == 'o')
+	      continue;
+	    break;
+	  case 'f':
+	    if (strncmp (save_decoded_options[j].orig_option_with_args_text
+			 + 2, "dump", 4) == 0)
+	      continue;
+	    break;
+	  default:
+	    break;
+	  }
+	VEC_safe_push (dchar_p, heap, switches,
+		       save_decoded_options[j].orig_option_with_args_text);
+	len += strlen (save_decoded_options[j].orig_option_with_args_text) + 1;
+	break;
+      }
+
+  producer = XNEWVEC (char, plen + 1 + len + 1);
+  tail = producer;
+  sprintf (tail, "%s %s", language_string, version_string);
+  tail += plen;
+
+  if (!dwarf_record_gcc_switches)
+    {
+#ifdef MIPS_DEBUGGING_INFO
+      /* The MIPS/SGI compilers place the 'cc' command line options in the
+	 producer string.  The SGI debugger looks for -g, -g1, -g2, or -g3;
+	 if they do not appear in the producer string, the debugger reaches
+	 the conclusion that the object file is stripped and has no debugging
+	 information.  To get the MIPS/SGI debugger to believe that there is
+	 debugging information in the object file, we add a -g to the producer
+	 string.  */
+      if (debug_info_level > DINFO_LEVEL_TERSE)
+	{
+	  memcpy (tail, " -g", 3);
+	  tail += 3;
+	}
+#endif
+    }
+
+  FOR_EACH_VEC_ELT (dchar_p, switches, j, p)
+    {
+      len = strlen (p);
+      *tail = ' ';
+      memcpy (tail + 1, p, len);
+      tail += len + 1;
+    }
+
+  *tail = '\0';
+  VEC_free (dchar_p, heap, switches);
+  return producer;
+}
+
 /* Generate the DIE for the compilation unit.  */
 
 static dw_die_ref
 gen_compile_unit_die (const char *filename)
 {
   dw_die_ref die;
-  char producer[250];
   const char *language_string = lang_hooks.name;
   int language;
 
@@ -18128,20 +18241,9 @@ gen_compile_unit_die (const char *filena
 	add_comp_dir_attribute (die);
     }
 
-  sprintf (producer, "%s %s", language_string, version_string);
-
-#ifdef MIPS_DEBUGGING_INFO
-  /* The MIPS/SGI compilers place the 'cc' command line options in the producer
-     string.  The SGI debugger looks for -g, -g1, -g2, or -g3; if they do
-     not appear in the producer string, the debugger reaches the conclusion
-     that the object file is stripped and has no debugging information.
-     To get the MIPS/SGI debugger to believe that there is debugging
-     information in the object file, we add a -g to the producer string.  */
-  if (debug_info_level > DINFO_LEVEL_TERSE)
-    strcat (producer, " -g");
-#endif
-
-  add_AT_string (die, DW_AT_producer, producer);
+  if (producer_string == NULL)
+    producer_string = gen_producer_string ();
+  add_AT_string (die, DW_AT_producer, producer_string);
 
   /* If our producer is LTO try to figure out a common language to use
      from the global list of translation units.  */
@@ -21774,6 +21876,15 @@ dwarf2out_finish (const char *filename)
   htab_t comdat_type_table;
   unsigned int i;
 
+  /* PCH might result in DW_AT_producer string being restored from the
+     header compilation, fix it up if needed.  */
+  dw_attr_ref producer = get_AT (comp_unit_die (), DW_AT_producer);
+  if (strcmp (AT_string (producer), producer_string) != 0)
+    {
+      struct indirect_string_node *node = find_AT_string (producer_string);
+      producer->dw_attr_val.v.val_str = node;
+    }
+
   gen_scheduled_generic_parms_dies ();
   gen_remaining_tmpl_value_param_die_attribute ();
 
--- gcc/Makefile.in.jj	2011-07-13 17:31:18.000000000 +0200
+++ gcc/Makefile.in	2011-07-14 10:36:40.000000000 +0200
@@ -2953,7 +2953,7 @@ dwarf2out.o : dwarf2out.c $(CONFIG_H) $(
    $(GGC_H) $(EXCEPT_H) dwarf2asm.h $(TM_P_H) langhooks.h $(HASHTAB_H) \
    gt-dwarf2out.h $(TARGET_H) $(CGRAPH_H) $(MD5_H) $(INPUT_H) $(FUNCTION_H) \
    $(GIMPLE_H) $(TREE_PASS_H) $(TREE_FLOW_H) $(CFGLAYOUT_H) \
-   tree-pretty-print.h $(COMMON_TARGET_H)
+   tree-pretty-print.h $(COMMON_TARGET_H) $(OPTS_H)
 dwarf2cfi.o : dwarf2cfi.c $(CONFIG_H) $(SYSTEM_H) coretypes.h $(TM_H) \
    version.h $(RTL_H) $(FUNCTION_H) $(DWARF2_H) dwarf2asm.h dwarf2out.h \
    $(GGC_H) $(TM_P_H) $(TARGET_H) $(TREE_PASS_H)


	Jakub

  reply	other threads:[~2011-07-14  9:42 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-07-13  0:17 Jakub Jelinek
2011-07-13 14:04 ` Jason Merrill
2011-07-13 14:34   ` Jakub Jelinek
2011-07-13 14:35     ` Jason Merrill
2011-07-14  9:48       ` Jakub Jelinek [this message]
2011-07-14 16:31         ` Jason Merrill
2011-07-14 16:32           ` Jakub Jelinek
2011-07-14 16:40             ` Jason Merrill
2011-07-21 20:22 ` Joseph S. Myers
2011-07-21 20:50   ` Jakub Jelinek
2011-07-21 21:01     ` Joseph S. Myers
2011-07-22 13:27       ` [PATCH] -grecord-gcc-switches (PR other/32998, take 2) Jakub Jelinek
2011-07-22 14:06         ` Joseph S. Myers
2011-07-22 14:11           ` [PATCH] -grecord-gcc-switches (PR other/32998, take 3) Jakub Jelinek
2011-07-22 20:43             ` Jason Merrill

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=20110714094148.GB2687@tyan-ft48-01.lab.bos.redhat.com \
    --to=jakub@redhat.com \
    --cc=gcc-patches@gcc.gnu.org \
    --cc=jason@redhat.com \
    --cc=jkratoch@redhat.com \
    --cc=joseph@codesourcery.com \
    --cc=roland@redhat.com \
    --cc=rth@redhat.com \
    --cc=tromey@redhat.com \
    /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).