Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
From: Mark Kettenis <mark.kettenis@xs4all.nl>
To: davem@davemloft.net
Cc: gdb-patches@sourceware.org
Subject: [PATCH] Fix complex float handling for 64-bit SPARC
Date: Sat, 07 Apr 2012 20:18:00 -0000	[thread overview]
Message-ID: <201204072017.q37KHeYH010978@glazunov.sibelius.xs4all.nl> (raw)

Hi Dave,

Looks like the 64-bit code you committed last year wasn't complete.
There is a missing sparc64_complex_float_p() check, which results in a
guarenteed assertion failure.  The diff below fixes this and as a
bonus also makes complex return values work.  Makes the complex float
tests in gdb.base/callfuncs.exp work with the OpenBSD system compiler
(GCC 4.2.1).  As far as I can tell this matches what's documented in
the Sun Studio 12 Fortran Programming Guide.

ok?


2012-04-07  Mark Kettenis  <kettenis@gnu.org>

	* sparc64-tdep.c (sparc64_store_arguments): Fix handling of
	complex floats, adjust some related comments and tighten a related
	assertion.
	(sparc64_extract_return_value): Handle complex floats.

Index: sparc64-tdep.c
===================================================================
RCS file: /cvs/src/src/gdb/sparc64-tdep.c,v
retrieving revision 1.57
diff -u -p -r1.57 sparc64-tdep.c
--- sparc64-tdep.c	4 Jan 2012 08:17:12 -0000	1.57
+++ sparc64-tdep.c	7 Apr 2012 20:12:08 -0000
@@ -785,7 +785,8 @@ sparc64_store_arguments (struct regcache
       struct type *type = value_type (args[i]);
       int len = TYPE_LENGTH (type);
 
-      if (sparc64_structure_or_union_p (type))
+      if (sparc64_structure_or_union_p (type) ||
+	  (sparc64_complex_floating_p (type) && len == 32))
 	{
 	  /* Structure or Union arguments.  */
 	  if (len <= 16)
@@ -816,10 +817,9 @@ sparc64_store_arguments (struct regcache
 	      num_elements++;
 	    }
 	}
-      else if (sparc64_floating_p (type))
+      else if (sparc64_floating_p (type) || sparc64_complex_floating_p (type))
 	{
 	  /* Floating arguments.  */
-
 	  if (len == 16)
 	    {
 	      /* The psABI says that "Each quad-precision parameter
@@ -887,7 +887,8 @@ sparc64_store_arguments (struct regcache
       int regnum = -1;
       gdb_byte buf[16];
 
-      if (sparc64_structure_or_union_p (type))
+      if (sparc64_structure_or_union_p (type) ||
+	  (sparc64_complex_floating_p (type) && len == 32))
 	{
 	  /* Structure or Union arguments.  */
 	  gdb_assert (len <= 16);
@@ -927,7 +928,7 @@ sparc64_store_arguments (struct regcache
 	      /* The psABI says "Each single-precision parameter value
                  will be assigned to one extended word in the
                  parameter array, and right-justified within that
-                 word; the left half (even floatregister) is
+                 word; the left half (even float register) is
                  undefined."  Even though the psABI says that "the
                  left half is undefined", set it to zero here.  */
 	      memset (buf, 0, 4);
@@ -960,7 +961,7 @@ sparc64_store_arguments (struct regcache
 	    }
 	  else if (regnum >= SPARC64_Q0_REGNUM && regnum <= SPARC64_Q8_REGNUM)
 	    {
-	      gdb_assert (element < 6);
+	      gdb_assert (element < 5);
 	      regnum = SPARC_O0_REGNUM + element;
 	      regcache_cooked_write (regcache, regnum, valbuf);
 	      regcache_cooked_write (regcache, regnum + 1, valbuf + 8);
@@ -1034,7 +1035,7 @@ sparc64_extract_return_value (struct typ
 	sparc64_extract_floating_fields (regcache, type, buf, 0);
       memcpy (valbuf, buf, len);
     }
-  else if (sparc64_floating_p (type))
+  else if (sparc64_floating_p (type) || sparc64_complex_floating_p (type))
     {
       /* Floating return values.  */
       for (i = 0; i < len / 4; i++)


             reply	other threads:[~2012-04-07 20:18 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-04-07 20:18 Mark Kettenis [this message]
2012-04-07 20:47 ` David Miller
2012-04-08  5:59   ` Mark Kettenis

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=201204072017.q37KHeYH010978@glazunov.sibelius.xs4all.nl \
    --to=mark.kettenis@xs4all.nl \
    --cc=davem@davemloft.net \
    --cc=gdb-patches@sourceware.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