issue with eraseall

issue with eraseall

[Earl Manning]

I am using the MTD stuff with a Samsung NAND chip and all is going well, except that I had to change a line in the eraseall program.  As it turns out the clean marker written by the eraseall program puts sizeof( struct jffs2_unknown_node ) [12 bytes] into the totlen element.  When jffs2 tries to mount the now blank partition, the cleanmarker check fails because the value is not 8.

In my implementation I just changed the value written by eraseall to be 8.  

The offending pieces of code are:

	util/eraseall.c:84
	fs/jffs2/wbuf.c:671-675

Is there a better place to post this information?

Thanks,
Earl Manning
 

Re: issue with eraseall

[Thomas Gleixner]

On Wednesday 02 April 2003 17:55, Earl Manning wrote:
> In my implementation I just changed the value written by eraseall to be 8.
> The offending pieces of code are:
>
> 	util/eraseall.c:84
> 	fs/jffs2/wbuf.c:671-675
I don't see a piece of offending code there

util/eraseall.c: 106 ...
		// format for JFFS2
		if (!jffs2) 
			continue;
				
		// write cleanmarker	
--->		if (isNAND) {
			struct mtd_oob_buf oob;
			oob.ptr = (unsigned char *) &cleanmarker;
			oob.start = erase.start;
--->			oob.start += meminfo.oobsize == 16 ? 8 : 6;
--->			oob.length = meminfo.oobsize == 16 ? 8 : 2;
			if (ioctl (fd, MEMWRITEOOB, &oob) != 0) {
				fprintf(stderr, "\n%s: %s: MTD writeoob failure: %s\n", exe_name, 
mtd_device, strerror(errno));
				continue;
			}
		} else {
This is for NOR !
		}


-- 
Thomas
________________________________________________________________________
linutronix - competence in embedded & realtime linux
http://www.linutronix.de
mail: tglx@linutronix.de