Re: [PATCH 4/5] ARM: gemini: add device tree for edimax NS2502

[LABBE Corentin]

Le Wed, May 19, 2021 at 01:17:59AM +0200, Linus Walleij a écrit :
> On Tue, May 18, 2021 at 11:55 AM LABBE Corentin <clabbe@xxxxxxxxxxxx> wrote:
> > Le Tue, May 18, 2021 at 01:46:20AM +0200, Linus Walleij a écrit :
> 
> > > Is the FIS broken since you hardcode the partitions?
> > >
> > > Doesn't this work:
> > >
> > > partitions {
> > >     compatible = "redboot-fis";
> > >     /* Eraseblock at 0x7e0000 */
> > >     fis-index-block = <0xfc>;
> > > };
> > >
> > > (Needs CONFIG_MTD_REDBOOT_PARTS)
> >
> > No it does not.
> >
> > physmap-flash 30000000.flash: no enabled pin control state
> > physmap-flash 30000000.flash: no disabled pin control state
> > physmap-flash 30000000.flash: initialized Gemini-specific physmap control
> > physmap-flash 30000000.flash: physmap platform flash device: [mem 0x30000000-0x331fffff]
> > 30000000.flash: Found 1 x16 devices at 0x0 in 16-bit bank. Manufacturer ID 0x000001 Chip ID 0x002201
> > number of CFI chips: 1
> > Searching for RedBoot partition table in 30000000.flash at offset 0x1fe0000
> > No RedBoot partition table detected in 30000000.flash
> > Searching for RedBoot partition table in 30000000.flash at offset 0x1fe0000
> > No RedBoot partition table detected in 30000000.flash
> 
> Hm since it searches at 0x1fe0000 that's where it thinks the last
> eraseblock is so the erase blocks are 0x20000 (128KB).
> 
> FIS at 0x7e0000 should be erase block
> 0x7e0000/0x20000 = 0x3f
> 
> Can you test fis-index-block = <0x3f> and see what happens?
> 

With 0x7F it works.

When testing, I found that fis-index-block was ignored. I just sent a patch fixing this.

Regards