This is the mail archive of the gdb@sources.redhat.com mailing list for the GDB project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

store.exp failures


When I run gdb.base/store.exp, (GCC 3.1, i686 Linux) I get tons and
tons of failures.  Looking into the log file, I see the following:

(gdb) break main
Breakpoint 1 at 0x80488f0: file gdb.base/store.c, line 233.
(gdb) run 
Starting program: /extra/gdb/mirror/src/gdb/testsuite/gdb.base/store 

Breakpoint 1, main () at gdb.base/store.c:233
233	  wack_struct_1 ();
(gdb) tbreak wack_char
Breakpoint 2 at 0x804849e: file gdb.base/store.c, line 46.
(gdb) PASS: gdb.base/store.exp: tbreak wack_char
continue
Continuing.

Program exited normally.

So, basically, it looks like some of the calls to the wack_XXX
functions are getting optimized out by the compiler, even though no
optimization flags are being passed.  Digging into the assembly
confirms this; the code for main starts off as follows:

main:
.LFB25:
	# store.c:215
	.loc 1 215 0
	# basic block 0
	pushl	%ebp
.LCFI67:
	movl	%esp, %ebp
.LCFI68:
	subl	$120, %esp
.LCFI69:
	andl	$-16, %esp
	movl	$0, %eax
	subl	%eax, %esp
	# store.c:233
	.loc 1 233 0
	leal	-10(%ebp), %eax
	movl	%eax, (%esp)
	call	wack_struct_1

I have no idea what to do about this.  The same thing happens with GCC
3.2.

For what it's worth, if I run it with GCC 2.95.3, I get FAILs on "new
up struct 1" and "new up struct 2", but everything else works.

(Incidentally, is 'wack' a typo for 'whack', or something else?)

David Carlton
carlton@math.stanford.edu


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]