This is the mail archive of the
newlib@sourceware.org
mailing list for the newlib project.
Re: gmtime_r hanging
- From: Joel Sherrill <joel dot sherrill at oarcorp dot com>
- To: Daniel O'Connor <doconnor at gsoft dot com dot au>, "newlib at sourceware dot org" <newlib at sourceware dot org>
- Date: Mon, 19 Dec 2011 08:21:15 -0500
- Subject: Re: gmtime_r hanging
There is always the possibility of a real bug but seeing a 128 byte buffer on the stack makes me wonder if you have blown the stack.
What is the stack address range? Check the stack pointer. You can always fill the stack memory with a pattern and see how much is written to.
--joel sherrill
RTEMS
Daniel O'Connor <doconnor@gsoft.com.au> wrote:
>Hi,
>I'm trying to use newlib on an STM32 and I ran across an oddity with gmtime_r
>
><snip>
> char buf[128];
> struct tm nowtm;
> time_t now;
><snip>
> now = time(NULL);
> printf("Time is %d\r\n", (int)now);
> gmtime_r(&now, &nowtm);
> strftime(buf, sizeof(buf) - 1, "Time is %Y/%m/%d %H:%M:%S\r\rn", &nowtm);
><snip>
>
>Breakpoint 1, gmtime_r (tim_p=0x2000ffdc, res=0x2000ffb8) at ../../../../../../gcc-4.6.1/newlib/libc/time/gmtime_r.c:13
>13 return (_mktm_r (tim_p, res, 1));
>(gdb) stepi
>0x0800175c 13 return (_mktm_r (tim_p, res, 1));
>(gdb)
>_mktm_r (tim_p=<value optimized out>, res=<value optimized out>, is_gmtime=<value optimized out>)
> at ../../../../../../gcc-4.6.1/newlib/libc/time/mktm_r.c:32
>32 {
>(gdb)
>0x08001fd2 32 {
>(gdb)
>0x08001fd4 32 {
>(gdb)
>0x08001fd6 32 {
>(gdb)
>0x08001fd8 32 {
>(gdb)
>0x08001fda 32 {
>(gdb)
>32 {
>(gdb)
>0x08001fde 32 {
>(gdb)
>0x08001fe0 32 {
>(gdb)
>38 __tzinfo_type *tz = __gettzinfo ();
>(gdb)
>__gettzinfo () at /Users/darius/projects/stm32test/syscalls.c:161
>161 }
>(gdb)
>0x080005c4 161 }
>(gdb)
>0x08001fe6 in _mktm_r (tim_p=0x2000ffdc, res=0x2000ffb8, is_gmtime=1) at ../../../../../../gcc-4.6.1/newlib/libc/time/mktm_r.c:38
>38 __tzinfo_type *tz = __gettzinfo ();
>(gdb)
>41 lcltime = *tim_p;
>(gdb)
>0x08001fea 41 lcltime = *tim_p;
>(gdb)
>43 days = ((long)lcltime) / SECSPERDAY;
>(gdb)
>0x08001fee 43 days = ((long)lcltime) / SECSPERDAY;
>(gdb)
>0x08001ff0 43 days = ((long)lcltime) / SECSPERDAY;
>(gdb)
>0x00000000 in ?? ()
>(gdb)
>
>I have no idea why it's jumping to NULL doing a division :(
>
>--
>Daniel O'Connor software and network engineer
>for Genesis Software - http://www.gsoft.com.au
>"The nice thing about standards is that there
>are so many of them to choose from."
> -- Andrew Tanenbaum
>GPG Fingerprint - 5596 B766 97C0 0E94 4347 295E E593 DC20 7B3F CE8C
>
>
>
>
>
>