Re: [LTP] [PATCH v3] linkat/linkat02.c: add new error number tests

[chrubis@suse.cz]

Hi!
> > +
> > +	SAFE_TOUCH(cleanup, TEST_EMLINK, 0666, NULL);
> > +
> > +	while (1) {
> > +		sprintf(lname, "%s%ld", BASENAME, ++link_max);
> > +		TEST(link(TEST_EMLINK, lname));
> > +		if (TEST_RETURN == -1) {
> > +			switch (TEST_ERRNO) {
> > +			case EMLINK:
> > +				tst_resm(TINFO, "for %s the max links is %ld",
> > +					 fstype, link_max);
> > +				break;
> > +			default:
> > +				tst_brkm(TBROK | TTERRNO, cleanup,
> > +					 "Unexpected error: ");
> > +				break;
> > +			}
> > +			break;
> > +		}
> > +	}
> > +}
> >
> Should it give a limit max number here when test EMLINK to handler the 
> filesystems which unlimited link count.
> AFAIK,it use 2^31 - 1 with xfs:
> /*
>   * The 32 bit link count in the inode theoretically maxes out at UINT_MAX.
>   * Since the pathconf interface is signed, we use 2^31 - 1 instead.
>   * The old inode format had a 16 bit link count, so its maximum is 
> USHRT_MAX.
>   */
> #define XFS_MAXLINK             ((1U << 31) - 1U)
> #define XFS_MAXLINK_1           65535U
> 
> We are not able to reach such limitation.

If xfs has no limit or not reachable limit on number of links we should
skip the setup for EMLINK and return TCONF in the EMLINK test if xfs
filesystem is used.

-- 
Cyril Hrubis
chrubis@suse.cz

------------------------------------------------------------------------------
Is your legacy SCM system holding you back? Join Perforce May 7 to find out:
&#149; 3 signs your SCM is hindering your productivity
&#149; Requirements for releasing software faster
&#149; Expert tips and advice for migrating your SCM now
http://p.sf.net/sfu/perforce
_______________________________________________
Ltp-list mailing list
Ltp-list@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ltp-list