On Thu, Sep 19, 2024 at 06:14:06PM +0300, Andy Shevchenko wrote:
> On Thu, Sep 19, 2024 at 04:58:14PM +0300, Serge Semin wrote:
> > The recently submitted fix-commit revealed a problem in the iDMA32
> > platform code. Even though the controller supported only a single master
> > the dw_dma_acpi_filter() method hard-coded two master interfaces with IDs
> > 0 and 1. As a result the sanity check implemented in the commit
> > b336268dde75 ("dmaengine: dw: Add peripheral bus width verification") got
> > incorrect interface data width and thus prevented the client drivers
> > from configuring the DMA-channel with the EINVAL error returned. E.g. the
> > next error was printed for the PXA2xx SPI controller driver trying to
> > configure the requested channels:
> >
> > > [ 164.525604] pxa2xx_spi_pci 0000:00:07.1: DMA slave config failed
> > > [ 164.536105] pxa2xx_spi_pci 0000:00:07.1: failed to get DMA TX descriptor
> > > [ 164.543213] spidev spi-SPT0001:00: SPI transfer failed: -16
> >
> > The problem would have been spotted much earlier if the iDMA32 controller
> > supported more than one master interfaces. But since it supports just a
> > single master and the iDMA32-specific code just ignores the master IDs in
> > the CTLLO preparation method, the issue has been gone unnoticed so far.
> >
> > Fix the problem by specifying a single master ID for both memory and
> > peripheral devices on the ACPI-based platforms if there is only one master
> > available on the controller. Thus the issue noticed for the iDMA32
> > controllers will be eliminated and the ACPI-probed DW DMA controllers will
> > be configured with the correct master ID by default.
>
> ...
>
> > static bool dw_dma_acpi_filter(struct dma_chan *chan, void *param)
> > {
> > + struct dw_dma *dw = to_dw_dma(chan->device);
> > struct acpi_dma_spec *dma_spec = param;
> > struct dw_dma_slave slave = {
> > .dma_dev = dma_spec->dev,
> > .src_id = dma_spec->slave_id,
> > .dst_id = dma_spec->slave_id,
> > .m_master = 0,
> > - .p_master = 1,
>
> I would leave this line as is and it makes more consistent in my opinion with
> the below comments which starts with the words "Fallback to...".
Ok.
>
> > };
> >
> > + /*
> > + * Fallback to using a single interface for both memory and peripheral
> > + * device if there is only one master I/F supported (e.g. iDMA32)
> > + */
> > + if (dw->pdata->nr_masters == 0)
>
> Why '== 0' and not '== 1'? Or '>= 2' if you wish to be on the save side (however,
> that '== 0' case is not obvious to me — do we really have that IRL?).
I several times checked the patch and never noticed this obvious typo.
Indeed nr_masters is the actual number of masters. So the statement should
have been '== 1'.
>
> > + slave.p_master = 0;
> > + else
> > + slave.p_master = 1;
>
> > +
> > +
>
> One blank line is enough.
Fully agreed.
I guess I was too hurrying to submit the fix so missed two stupid
mistakes in just 7-lines patch. "Nice" anti-record for me. Sorry about
that and much appreciated for reviewing the bit. I'll resubmit v2
shortly.
-Serge(y)
>
> > return dw_dma_filter(chan, &slave);
> > }
>
> ...
>
> P.S. I'll test it later this or next week, if Ferry wouldn't beat me up to it.
>
> --
> With Best Regards,
> Andy Shevchenko
>
>
