[U-Boot] [0/6] Patch serie UBI min I/O size fixes

[U-Boot] [0/6] Patch serie UBI min I/O size fixes

[Holger Brunck]

Hi,

Brunck, Holger wrote:
> This patch serie fixes a problem in the UBI layer. The min I/O size has
> currently a fixed value of 1 for NOR flashes. But for NOR flashes the min I/O
> size has to be equal to the size of the writebuffer of the flash. Therefore the
> mtd layer was enhanced with the writebuffer which was later on used in the ubi
> layer to set the min I/O size.
> 
> This patches fixes problems with flash types with a writebuffersize > 128 Byte
> and is mainly taken from the similar linux patches. See MTD mailing list posts:
> http://lists.infradead.org/pipermail/linux-mtd/2010-December/033460.html
> 
> Without the u-boot patches attaching to a ubi device with a min I/O buffer
> larger than 128 would fail.
> 

with the same changes in the linux kernel we see incompatibilities for the
UBIFS. Due to this fact the patches should be restrained until this problem is
solved.

See:
http://lists.infradead.org/pipermail/linux-mtd/2011-January/033794.html

Best regards
Holger Brunck

[U-Boot] [0/6] Patch serie UBI min I/O size fixes

[Detlev Zundel]

Hi Holger,

> Hi,
>
> Brunck, Holger wrote:
>> This patch serie fixes a problem in the UBI layer. The min I/O size has
>> currently a fixed value of 1 for NOR flashes. But for NOR flashes the min I/O
>> size has to be equal to the size of the writebuffer of the flash. Therefore the
>> mtd layer was enhanced with the writebuffer which was later on used in the ubi
>> layer to set the min I/O size.
>> 
>> This patches fixes problems with flash types with a writebuffersize > 128 Byte
>> and is mainly taken from the similar linux patches. See MTD mailing list posts:
>> http://lists.infradead.org/pipermail/linux-mtd/2010-December/033460.html
>> 
>> Without the u-boot patches attaching to a ubi device with a min I/O buffer
>> larger than 128 would fail.
>> 
>
> with the same changes in the linux kernel we see incompatibilities for the
> UBIFS. Due to this fact the patches should be restrained until this problem is
> solved.
>
> See:
> http://lists.infradead.org/pipermail/linux-mtd/2011-January/033794.html

Ah, I see.  Thanks for the info.

In defense of my overlapping mail, I can only state that the reference
header in this mail of yours does not point to the earlier thread (check
[1]) and thus I saw it only much later:

> References: <4D41331C.7000302@keymile.com>
> In-Reply-To: <4D41331C.7000302@keymile.com>

But anyway, sorry for the noise.

Cheers
  Detlev

[1] http://mid.gmane.org/4D41331C.7000302 at keymile.com

-- 
Man against god... God against Man... Man against nature... Nature against
man...   God against nature...   Nature against god...  a very, very funny
religion.             -- Zen Master D.T. Suzuki commenting on Christianity
--
DENX Software Engineering GmbH,      MD: Wolfgang Denk & Detlev Zundel
HRB 165235 Munich,  Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: (+49)-8142-66989-40 Fax: (+49)-8142-66989-80 Email: dzu at denx.de

[U-Boot] [0/6] Patch serie UBI min I/O size fixes

[Holger Brunck]

This patch serie fixes a problem in the UBI layer. The min I/O size has
currently a fixed value of 1 for NOR flashes. But for NOR flashes the min I/O
size has to be equal to the size of the writebuffer of the flash. Therefore the
mtd layer was enhanced with the writebuffer which was later on used in the ubi
layer to set the min I/O size.

This patches fixes problems with flash types with a writebuffersize > 128 Byte
and is mainly taken from the similar linux patches. See MTD mailing list posts:
http://lists.infradead.org/pipermail/linux-mtd/2010-December/033460.html

Without the u-boot patches attaching to a ubi device with a min I/O buffer
larger than 128 would fail.

Best
Regards

[U-Boot] [0/6] Patch serie UBI min I/O size fixes

[Stefan Roese]

Hi Holger,

On Thursday 20 January 2011 17:16:32 Holger Brunck wrote:
> This patch serie fixes a problem in the UBI layer. The min I/O size has
> currently a fixed value of 1 for NOR flashes. But for NOR flashes the min
> I/O size has to be equal to the size of the writebuffer of the flash.
> Therefore the mtd layer was enhanced with the writebuffer which was later
> on used in the ubi layer to set the min I/O size.
> 
> This patches fixes problems with flash types with a writebuffersize > 128
> Byte and is mainly taken from the similar linux patches. See MTD mailing
> list posts:
> http://lists.infradead.org/pipermail/linux-mtd/2010-December/033460.html
> 
> Without the u-boot patches attaching to a ubi device with a min I/O buffer
> larger than 128 would fail.

IIRC (please correct me if I'm wrong), the Linux part of these patches has 
undergone some changes. Is this U-Boot patch series still up-to-date?

Cheers,
Stefan

--
DENX Software Engineering GmbH,      MD: Wolfgang Denk & Detlev Zundel
HRB 165235 Munich,  Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: (+49)-8142-66989-0 Fax: (+49)-8142-66989-80 Email: office at denx.de

[U-Boot] [0/6] Patch serie UBI min I/O size fixes

[Holger Brunck]

Hi Stefan,

Stefan Roese wrote:
> Hi Holger,
> 
> On Thursday 20 January 2011 17:16:32 Holger Brunck wrote:
>> This patch serie fixes a problem in the UBI layer. The min I/O size has
>> currently a fixed value of 1 for NOR flashes. But for NOR flashes the min
>> I/O size has to be equal to the size of the writebuffer of the flash.
>> Therefore the mtd layer was enhanced with the writebuffer which was later
>> on used in the ubi layer to set the min I/O size.
>>
>> This patches fixes problems with flash types with a writebuffersize > 128
>> Byte and is mainly taken from the similar linux patches. See MTD mailing
>> list posts:
>> http://lists.infradead.org/pipermail/linux-mtd/2010-December/033460.html
>>
>> Without the u-boot patches attaching to a ubi device with a min I/O buffer
>> larger than 128 would fail.
> 
> IIRC (please correct me if I'm wrong), the Linux part of these patches has 
> undergone some changes. Is this U-Boot patch series still up-to-date?
> 

no. The assumption that (mtd->writebufsize == min I/O size in UBI) for NOR
flashes was incorrect and the  patch UBI: use mtd->writebufsize to set minimal
I/O unit size was reverted in the mainline kernel. Whats still part in the
kernel is the enhancement of the mtd struct with the writebufsize. But also in
this topic some changes are made. So the patch series should be dropped or
updated with the recent changes.

But currently due to the changes I did not see a usage for this additional value
for u-boot. For this reason I propose to drop the patch series.

Best regards
Holger