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

List:       busybox
Subject:    Re: [PATCHv4] nandwrite: new mtd-utils applet
From:       Baruch Siach <baruch () tkos ! co ! il>
Date:       2010-08-29 7:33:58
Message-ID: 20100829073358.GA8285 () jasper ! tkos ! co ! il
[Download RAW message or body]

Hi Denys,

On Wed, Aug 25, 2010 at 04:38:08PM +0200, Denys Vlasenko wrote:
> On Wed, Aug 25, 2010 at 6:38 AM, Baruch Siach <baruch@tkos.co.il> wrote:
> > + static unsigned next_good_eraseblock(int fd, struct mtd_info_user *meminfo,
> > +               unsigned block_offset)
> > +{
> > +       loff_t offs = block_offset;
> > +
> > +       while (1) {
> > +               if (block_offset >= meminfo->size)
> > +                       bb_error_msg_and_die("not enough space in MTD device");
> > +               if (xioctl(fd, MEMGETBADBLOCK, &offs)) {
> > +                       offs += meminfo->erasesize;
> > +                       printf("Skipping bad block at 0x%08llx\n", offs);
> 
> The message prints wrong offset (offs +=... should be after the message).
> %llx does not necessarily match loff_t type.
> You forgot to increment block_offset, thus you can run off the MTD.
> 
> > +               } else
> > +                       return offs;
> > +       }
> > +}
> 
> How about this?
> 
>         while (1) {
>                 loff_t offs;
>                 if (block_offset >= meminfo->size)
>                         bb_error_msg_and_die("not enough space in MTD device");
>                 offs = block_offset;
>                 if (xioctl(fd, MEMGETBADBLOCK, &offs) == 0)
>                         return block_offset;
>                 /* ioctl returned 1 = "bad block" */
>                 printf("Skipping bad block at 0x%08x\n", block_offset);
>                 block_offset += meminfo->erasesize;
>         }
> 
> Applied with a few changes:
> 
> http://git.busybox.net/busybox/commit/?id=6f32ea4039535c48759a217fd6352193846a393c

Thanks.

> Please test current git.

I've found two problems:

1. The move from bb_strtou() to xatou() breaks hex offsets

2. The first erase block isn't checked when mtdoffset is not erase block 
   aligned

I'll post two patches fixing these problems as a reply to this message 
shortly.

baruch

-- 
                                                     ~. .~   Tk Open Systems
=}------------------------------------------------ooO--U--Ooo------------{=
   - baruch@tkos.co.il - tel: +972.2.679.5364, http://www.tkos.co.il -
_______________________________________________
busybox mailing list
busybox@busybox.net
http://lists.busybox.net/mailman/listinfo/busybox

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