[prev in list] [next in list] [prev in thread] [next in thread] 

List:       ast-users
Subject:    Re: [ast-users] strange ioctl32 messages on linux
From:       Glenn Fowler <gsf () research ! att ! com>
Date:       2010-03-07 5:02:05
Message-ID: 201003070502.o27525Gw017896 () penguin ! research ! att ! com
[Download RAW message or body]


On Sun, 7 Mar 2010 11:20:51 +0700 BuraphaLinux Server wrote:
>    It appears your guess was correct.  I tried to figure out how to do
> that, and came up with the attached patch (against
> ast-ksh.2009-05-05.tgz).  I'm sorry I could not understand how to do
> this using the build system.  It does remove the dmesg log spamming
> problem.  But is it a good solution?

the place to patch is
	src/lib/libast/features/lib
I'll change the stream_peek test to specifically check if I_PEEK works on pipes
---
tst	stream_peek note{ ioctl(I_PEEK) works on pipes }end execute{
	#include <sys/types.h>
	#include <unistd.h>
	#include <stropts.h>
	int
	main()
	{	struct strpeek	peek;
		int		fds[2];
		char		ctlbuf[32];
		char		databuf[32];
		peek.flags = 0;
		peek.ctlbuf.maxlen = peek.ctlbuf.len = sizeof(ctlbuf);
		peek.ctlbuf.buf = ctlbuf;
		peek.databuf.maxlen = peek.databuf.len = sizeof(databuf);
		peek.databuf.buf = databuf;
		pipe(fds);
		return ioctl(fds[0],I_PEEK,&peek) < 0;
	}
}end
---
>    Thank you for explaining the .exe and expected crashes.  I guess it
> is in the documentation, but I missed it.

the internals of the iffe/probe tests are not documented

_______________________________________________
ast-users mailing list
ast-users@research.att.com
https://mailman.research.att.com/mailman/listinfo/ast-users
[prev in list] [next in list] [prev in thread] [next in thread]