[commit] fix gdbserver build failure on amd64-linux

[commit] fix gdbserver build failure on amd64-linux

[Doug Evans]

Hi.

fyi, I committed this.

2012-07-02  Doug Evans  <dje@google.com>

	* mem-break.c (gdb_no_commands_at_breakpoint): Fix cast from
	pointer to int.

Index: mem-break.c
===================================================================
RCS file: /cvs/src/src/gdb/gdbserver/mem-break.c,v
retrieving revision 1.34
diff -u -p -r1.34 mem-break.c
--- mem-break.c	2 Jul 2012 15:29:38 -0000	1.34
+++ mem-break.c	2 Jul 2012 17:10:13 -0000
@@ -22,6 +22,7 @@
 #include "server.h"
 #include "regcache.h"
 #include "ax.h"
+#include <stdint.h>
 
 const unsigned char *breakpoint_data;
 int breakpoint_len;
@@ -929,8 +930,8 @@ gdb_no_commands_at_breakpoint (CORE_ADDR
     return 0;
 
   if (debug_threads)
-    fprintf (stderr, "at 0x%s, bp command_list is 0x%x\n",
-	     paddress (where), (int) bp->command_list);
+    fprintf (stderr, "at 0x%s, bp command_list is 0x%lx\n",
+	     paddress (where), (long) (uintptr_t) bp->command_list);
   return (bp->command_list == NULL);
 }
 

Re: [commit] fix gdbserver build failure on amd64-linux

[Pedro Alves]

On 07/02/2012 06:12 PM, Doug Evans wrote:
> Hi.
> 
> fyi, I committed this.
> 
> 2012-07-02  Doug Evans  <dje@google.com>
> 
> 	* mem-break.c (gdb_no_commands_at_breakpoint): Fix cast from
> 	pointer to int.
> 
> Index: mem-break.c
> ===================================================================
> RCS file: /cvs/src/src/gdb/gdbserver/mem-break.c,v
> retrieving revision 1.34
> diff -u -p -r1.34 mem-break.c
> --- mem-break.c	2 Jul 2012 15:29:38 -0000	1.34
> +++ mem-break.c	2 Jul 2012 17:10:13 -0000
> @@ -22,6 +22,7 @@
>  #include "server.h"
>  #include "regcache.h"
>  #include "ax.h"
> +#include <stdint.h>
>  
>  const unsigned char *breakpoint_data;
>  int breakpoint_len;
> @@ -929,8 +930,8 @@ gdb_no_commands_at_breakpoint (CORE_ADDR
>      return 0;
>  
>    if (debug_threads)
> -    fprintf (stderr, "at 0x%s, bp command_list is 0x%x\n",
> -	     paddress (where), (int) bp->command_list);
> +    fprintf (stderr, "at 0x%s, bp command_list is 0x%lx\n",
> +	     paddress (where), (long) (uintptr_t) bp->command_list);

That's not correct on e.g., Win64, where sizeof long < sizeof pointer.
The right fix is to either use host_address_to_string, or %p, if acceptable
on all hosts.

>    return (bp->command_list == NULL);
>  }
>  
> 



-- 
Pedro Alves

Re: [commit] fix gdbserver build failure on amd64-linux

[Tom Tromey]

Doug> -    fprintf (stderr, "at 0x%s, bp command_list is 0x%x\n",
Doug> -	     paddress (where), (int) bp->command_list);
Doug> +    fprintf (stderr, "at 0x%s, bp command_list is 0x%lx\n",
Doug> +	     paddress (where), (long) (uintptr_t) bp->command_list);

I think it is more normal to use host_address_to_string here.

Tom

Re: [commit] fix gdbserver build failure on amd64-linux

[Stan Shebs]

On 7/2/12 10:41 AM, Tom Tromey wrote:
> Doug> -    fprintf (stderr, "at 0x%s, bp command_list is 0x%x\n",
> Doug> -	     paddress (where), (int) bp->command_list);
> Doug> +    fprintf (stderr, "at 0x%s, bp command_list is 0x%lx\n",
> Doug> +	     paddress (where), (long) (uintptr_t) bp->command_list);
>
> I think it is more normal to use host_address_to_string here.
>
Or maybe dump the debug print altogether - I'm looking at it and 
wondering why I ever thought this one was going to be useful. :-)

Stan