Hello,
This is the fragment where it fails:
It seems that the call to regcomp is causing the issue.
I don't think this has anything to do with direct access to the hardware. It seems to be some kind of pointer issue, the call seems to access data it can't reach.
/*********************************************************************** ***** * compile_reg_expr * * Compile a regular expression.
************************************************************************ ****/ void compile_reg_expr(int cflags, const char *expr, regex_t *reg) { static const size_t ERROR_BUF_SIZE = 256; char error_msg[ERROR_BUF_SIZE]; int result;
-->>>> if ((result = regcomp(reg, expr, cflags)) != 0) { <<---------
regerror(result, reg, error_msg, ERROR_BUF_SIZE); fprintf(stderr, "%s: %s\n", prog_name, error_msg); exit(1); }
printf( "compile_reg_expr 3\n");
}
Regards,
Wim
-----Original Message----- From: coreboot-bounces@coreboot.org [mailto:coreboot-bounces@coreboot.org] On Behalf Of Patrick Georgi Sent: dinsdag 7 mei 2013 8:51 To: coreboot@coreboot.org Subject: Re: [coreboot] Building AMD Persimmon in MinGW
Am 2013-05-07 01:56, schrieb Peter Stuge:
Sounds right. Would be great to find out more details about why and how nvramtool is being used during the build!
During the handling of cmos.layout and cmos.settings (if present and used).
We don't need any special hardware access for that and should not use any such code while building coreboot. One place to look at is the win32mmap implementation, which _is_ used.
Patrick