Uboot SPI uclass - question

[Simon Glass]

Hi,

On Mon, 23 Nov 2020 at 00:22, Rasmus Villemoes
<rasmus.villemoes at prevas.dk> wrote:
>
> On 02/11/2020 19.14, Moshe, Yaniv wrote:
> > Hi Simon,
> > Hope you're doing good.
> >
> > My name is Yaniv, and I have a question regarding an old commit in U-boot:
> > commit 60e2809a848bccd3a8090d3f2237964670f2780c
> > Author: Simon Glass <sjg at chromium.org<mailto:sjg at chromium.org>>
> > Date:   Tue Feb 17 15:29:35 2015 -0700
> >
> >     dm: spi: Avoid setting the speed with every transfer
> >
> >     Only set the speed if it has changed from last time. Since the speed will
> >     be 0 when the device is probed it will always be changed on the first
> >     transfer after the device is probed.
> >
> >     Signed-off-by: Simon Glass sjg at chromium.org<mailto:sjg at chromium.org>
> >
> >
> > As far as I understand it, each SPI slave has its own struct. So, when we will use the device for the first time, we will set the speed.
> > But, if we're jumping between different slaves, the slaves structs were already configured at the first time, and since then there won't be any future speed setting.
> >
> > e.g:
> > access flash with max-freq = X - > speed set
> > access TPM with max-freq =Y -> speed set
> > access flash again, slave struct already configured, no change between speed and slave->speed, -> speed stay the same as for TPM.
> >
> > Did I got it wrong? I'm trying to understand what am I missing...
>
> No, you're reading it the same way I did:
>
> https://lists.denx.de/pipermail/u-boot/2019-December/393854.html
>
> I never followed up because I didn't actually have two different spi
> devices in play on the board I was working on at the time. But it's
> something that really should be fixed.

It is easy enough to add a sandbox test for this I think.

Regards,
Simon