Re: [Fdutils] DMA cache consistency bug introduced in 2.6.28

[alain <alain@knaff.lu>]

Linus Torvalds wrote:

> diff --git a/drivers/block/floppy.c b/drivers/block/floppy.c
> index 3266b4f..9c9148c 100644
> --- a/drivers/block/floppy.c
> +++ b/drivers/block/floppy.c
> @@ -2237,13 +2237,10 @@ static void setup_format_params(int track)
>  	for (count = 1; count <= F_SECT_PER_TRACK; ++count) {
>  		here[n].sect = count;
>  		n = (n + il) % F_SECT_PER_TRACK;
> -		if (here[n].sect) {	/* sector busy, find next free sector */
> +		while (here[n].sect) {	/* sector busy, find next free sector */
>  			++n;
> -			if (n >= F_SECT_PER_TRACK) {
> +			if (n >= F_SECT_PER_TRACK)
>  				n -= F_SECT_PER_TRACK;
> -				while (here[n].sect)
> -					++n;
> -			}
>  		}
>  	}
>  	if (_floppy->stretch & FD_SECTBASEMASK) {

The original code does indeed look a little bit strange... and might
break if there is a long run of "busy" sectors near the end of the
physical track. Or maybe there is a mathematical reason why this
situation cannot occur. I'll have to think about it a little bit more to
come up with a test case that will break either the new or old code.

But in any case, if a bug would occur due to this code, it would only
depend on the format's parameters, and not on the hardwarde.

Regards,

Alain