Date: Thu, 28 Jul 2005 09:59:28 +0200 From: Uwe Doering <gemini@geminix.org> To: Rink Springer <rink@stack.nl> Cc: hackers@freebsd.org, Ludvig Strigeus <strigeus@gmail.com>, ed@fxq.nl Subject: Re: Problem with pic16l_setled Message-ID: <42E89060.3060700@geminix.org> In-Reply-To: <20050727134432.GA10142@stack.nl> References: <4285cd45050727063146c054d4@mail.gmail.com> <20050727134432.GA10142@stack.nl>
next in thread | previous in thread | raw e-mail | index | archive | help
Rink Springer wrote:
> * Ludvig Strigeus (strigeus@gmail.com) wrote:
>
>>I see this code:
>>
>>+ENTRY(pic16l_setled)
>>+ push %ebp
>>+ mov %esp,%ebp
>>+
>>+ movl 0x8(%ebp),%ebx
>>+ orl $0x800,%ebx
>>+ call pic16l_ledhlp
>>+ movl $0x701,%ebx
>>+ call pic16l_ledhlp
>>+
>>+ leave
>>+ ret
>>
>>With the standard x86 calling convention, you're not supposed to
>>modify ebx, esi or edi without saving them first. Try adding a
>>push/pop ebx around.
>
> I doubt this is the problem (I've looked at other .s files like
> i386/i386/support.s, function ssdtosd, and they don't seem to do this).
Well, at least in 'support.s' of RELENG_4 register ebx _does_ get pushed
onto the stack in function 'ssdtosd'. I suppose this is true also for
the other FreeBSD branches.
If you call a function from the C level only registers eax, edx and ecx
may be altered without preserving their original contents on the stack
(or wherever else). That is, in C the calling level expects that the
other registers survive a function call unharmed. As pointed out that's
the standard x86 calling convention which applies to all x86 compilers,
not just GCC.
Uwe
--
Uwe Doering | EscapeBox - Managed On-Demand UNIX Servers
gemini@geminix.org | http://www.escapebox.net
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?42E89060.3060700>