[Ocfs2-tools-devel] [PATCH 1/1] Ocfs2-tools: Add new corruptions of inline-data for fswreck.

tristan.ye tristan.ye at oracle.com
Mon Sep 8 03:06:40 PDT 2008


On Mon, 2008-09-08 at 17:04 +0800, Tao Ma wrote:
> Hi Tristan,
> 	The patch looks great. Just some minor comments. ;)
> 
> Tristan Ye wrote:
> 
> > +void mess_up_inline_flag(ocfs2_filesys *fs, uint64_t blkno)
> > +{
> > +
> > +	int i;
> > +	errcode_t ret;
> > +	char *buf = NULL;
> > +	uint64_t inline_blkno;
> > +	struct ocfs2_dinode *di;
> > +
> > +	for (i = 0; i < 2; i++) {
> > +		if (i == 0)
> > +			create_file(fs, blkno, &inline_blkno);
> > +		else
> > +			create_directory(fs, blkno, &inline_blkno);
> > +
> > +		ret = ocfs2_malloc_block(fs->fs_io, &buf);
> > +		if (ret)
> > +			FSWRK_COM_FATAL(progname, ret);
> you can move this malloc out of "for" so that it will only be allocated 
> once while current it will be alloced and freed "i" times. The same goes 
> with mess_up_inline_count.
> And as for mess_up_inline_flag, it should be used to test an 
> old-formatted volume with a inlined file, so you'd better bail out if 
> this volume is already formatted with inline-data supported.
> 
> btw, could you please add some output so that user can know what's 
> happening. See how other corrupt code do as a reference.

that's really great catches,will do the corresponding modifications:)

Tristan.



> 
> Regards,
> Tao




More information about the Ocfs2-tools-devel mailing list