Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
From: Pedro Alves <palves@redhat.com>
To: gdb-patches@sourceware.org
Subject: [PATCH] [sim/ppc] Fix printf_filtered reference
Date: Tue, 17 Nov 2015 19:07:00 -0000	[thread overview]
Message-ID: <1447787243-13888-1-git-send-email-palves@redhat.com> (raw)

Building a gdb that includes the PPC sim in C++ mode fails to link with:

(...)s.o compile-object-load.o compile-object-run.o compile-loc2c.o compile-c-support.o inflow.o    init.o \
          ../sim/ppc/libsim.a ../readline/libreadline.a ../opcodes/libopcodes.a ../bfd/libbfd.a -lz  ../libiberty/libiberty.a ../libdecnumber/libdecnumber.a    -ldl -ldl -lncurses -lm -ldl  -lguile-2.0 -lgc  -lpthread -ldl -lutil -lm -lpython2.7 -Xlinker -export-dynamic -lexpat -llzma -lbabeltrace -lbabeltrace-ctf  ../libiberty/libiberty.a  build-gnulib/import/libgnu.a
../sim/ppc/libsim.a(sim_calls.o): In function `sim_open':
/home/pedro/gdb/mygit/cxx-convertion/src/sim/ppc/sim_calls.c:73: undefined reference to `printf_filtered'
/home/pedro/gdb/mygit/cxx-convertion/src/sim/ppc/sim_calls.c:73: undefined reference to `printf_filtered'
../sim/ppc/libsim.a(sim_calls.o): In function `sim_close':
/home/pedro/gdb/mygit/cxx-convertion/src/sim/ppc/sim_calls.c:93: undefined reference to `printf_filtered'
/home/pedro/gdb/mygit/cxx-convertion/src/sim/ppc/sim_calls.c:93: undefined reference to `printf_filtered'
../sim/ppc/libsim.a(sim_calls.o): In function `sim_load':
/home/pedro/gdb/mygit/cxx-convertion/src/sim/ppc/sim_calls.c:102: undefined reference to `printf_filtered'
../sim/ppc/libsim.a(sim_calls.o):/home/pedro/gdb/mygit/cxx-convertion/src/sim/ppc/sim_calls.c:102: more undefined references to `printf_filtered' follow
collect2: error: ld returned 1 exit status

The undefined references come from TRACE macro calls, which expand to
calls to printf_filtered.

But note that the sim's 'printf_filtered' is actually a #define to
'sim_io_printf_filtered', in sim_callbacks.h :

 #define printf_filtered sim_io_printf_filtered

AFAICS, this is not meant to call gdb's printf_filtered function.  The
ChangeLog entry that added the printf_filtered macro reads:

 Tue Jul 30 21:12:24 1996  Andrew Cagney  <cagney@kremvax.highland.com.au>

	 * sim_callbacks.h (sim_io_printf_filtered): Replace
	 printf_filtered with a local simulator specific version.  Add
	 #define printf_filtered to simplify updating of existing code.

That is, just another incomplete/partial transition.  Maybe prior to
1996 this was really meant to call gdb's printf_filtered version.

The reference to printf_filtered appears because sim_calls.c, the
compilation unit that fails to link, has this at the top:

 #undef printf_filtered /* blow away the mapping */

presumably so that this further below:

 void
 sim_io_printf_filtered(const char *fmt,
			...)
 {
   (...)
   callbacks->printf_filtered(callbacks, "%s", message);
 }

works.  So those TRACE macros instances in sim_calls.c just happen to
work because gdb is linked in, which satisfies the 'printf_filtered'
reference, when GDB is built in C mode.  When built in C++ mode, the
problem is exposed, as GDB's printf_filtered is mangled.

The fix here is to make the TRACE macro call sim_io_printf_filtered
directly.

(Standalone "run" doesn't fail to link simply because the offending
routines are not part of its link.)

sim/ppc/ChangeLog
2015-11-17  Pedro Alves  <palves@redhat.com>

	* debug.h (TRACE, ITRACE, DTRACE, DITRACE, PTRACE): Call
	sim_io_printf_filtered instead of printf_filtered.
---
 sim/ppc/debug.h | 32 ++++++++++++++++----------------
 1 file changed, 16 insertions(+), 16 deletions(-)

diff --git a/sim/ppc/debug.h b/sim/ppc/debug.h
index fdf2e3c..6f19624 100644
--- a/sim/ppc/debug.h
+++ b/sim/ppc/debug.h
@@ -87,8 +87,8 @@ extern int ppc_trace[nr_trace_options];
 do { \
   if (WITH_TRACE) { \
     if (ppc_trace[OBJECT]) { \
-      printf_filtered("%s:%d: ", filter_filename(__FILE__), __LINE__); \
-      printf_filtered ARGS; \
+      sim_io_printf_filtered("%s:%d: ", filter_filename(__FILE__), __LINE__); \
+      sim_io_printf_filtered ARGS; \
     } \
   } \
 } while (0)
@@ -98,8 +98,8 @@ do { \
 do { \
   if (WITH_TRACE) { \
     if (ppc_trace[OBJECT]) { \
-      printf_filtered("%s:%d:0x%08lx:%s ", itable[MY_INDEX].file, itable[MY_INDEX].line_nr, (long)cia, itable[MY_INDEX].name); \
-      printf_filtered ARGS; \
+      sim_io_printf_filtered("%s:%d:0x%08lx:%s ", itable[MY_INDEX].file, itable[MY_INDEX].line_nr, (long)cia, itable[MY_INDEX].name); \
+      sim_io_printf_filtered ARGS; \
     } \
   } \
 } while (0)
@@ -112,11 +112,11 @@ do { \
     if (ppc_trace[trace_devices] \
 	|| ppc_trace[trace_##OBJECT##_device] \
 	|| trace_device) { \
-      printf_filtered("%s:%d:%s:%s%s ", \
-		      filter_filename(__FILE__), __LINE__, #OBJECT, \
-		      trace_device ? device_path(me) : "", \
-		      trace_device ? ":" : ""); \
-      printf_filtered ARGS; \
+      sim_io_printf_filtered("%s:%d:%s:%s%s ",					\
+			     filter_filename(__FILE__), __LINE__, #OBJECT, \
+			     trace_device ? device_path(me) : "",	\
+			     trace_device ? ":" : "");			\
+      sim_io_printf_filtered ARGS; \
     } \
   } \
 } while (0)
@@ -130,11 +130,11 @@ do { \
     if (ppc_trace[trace_devices] \
 	|| ppc_trace[trace_##OBJECT##_device] \
 	|| trace_device) { \
-      printf_filtered("%s:%d:%s:%s%s ", \
-		      filter_filename(__FILE__), __LINE__, #OBJECT, \
-		      trace_device ? device_path(me) : "", \
-		      trace_device ? ":" : ""); \
-      printf_filtered ARGS; \
+      sim_io_printf_filtered("%s:%d:%s:%s%s ", \
+			     filter_filename(__FILE__), __LINE__, #OBJECT, \
+			     trace_device ? device_path(me) : "",	\
+			     trace_device ? ":" : "");			\
+      sim_io_printf_filtered ARGS; \
     } \
   } \
 } while (0)
@@ -144,8 +144,8 @@ do { \
 do { \
   if (WITH_TRACE) { \
     if (ppc_trace[trace_##OBJECT##_package]) { \
-      printf_filtered("%s:%d:%s: ", filter_filename(__FILE__), __LINE__, #OBJECT); \
-      printf_filtered ARGS; \
+      sim_io_printf_filtered("%s:%d:%s: ", filter_filename(__FILE__), __LINE__, #OBJECT); \
+      sim_io_printf_filtered ARGS; \
     } \
   } \
 } while (0)
-- 
1.9.3


             reply	other threads:[~2015-11-17 19:07 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-11-17 19:07 Pedro Alves [this message]
2015-11-17 19:15 ` Mike Frysinger
2015-11-17 19:25   ` Pedro Alves

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=1447787243-13888-1-git-send-email-palves@redhat.com \
    --to=palves@redhat.com \
    --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