RE: [PATCH v2 3/3] dmaengine: tegra: Add support for dma-channel-mask

From: Akhil R
Date: Fri Sep 23 2022 - 07:09:26 EST


> On 23/09/2022 11:17, Akhil R wrote:
> >> On 19/09/2022 12:25, Akhil R wrote:
> >>> Add support for dma-channel-mask so that only the specified channels
> >>> are used. This helps to reserve some channels for the firmware.
> >>>
> >>> This was initially achieved by limiting the channel number to 31 in
> >>> the driver and adjusting the register address to skip channel0 which
> >>> was reserved for a firmware. Now, with this change, the driver can
> >>> align more to the actual hardware which has 32 channels.
> >>>
> >>> Signed-off-by: Akhil R <akhilrajeev@xxxxxxxxxx>
> >>> ---
> >>> drivers/dma/tegra186-gpc-dma.c | 37 +++++++++++++++++++++++++++---
> ----
> >>> 1 file changed, 30 insertions(+), 7 deletions(-)
> >>>
> >>> diff --git a/drivers/dma/tegra186-gpc-dma.c b/drivers/dma/tegra186-gpc-
> >> dma.c
> >>> index fa9bda4a2bc6..1d1180db6d4e 100644
> >>> --- a/drivers/dma/tegra186-gpc-dma.c
> >>> +++ b/drivers/dma/tegra186-gpc-dma.c
> >>> @@ -161,7 +161,10 @@
> >>> #define TEGRA_GPCDMA_BURST_COMPLETION_TIMEOUT 5000 /* 5
> >> msec */
> >>>
> >>> /* Channel base address offset from GPCDMA base address */
> >>> -#define TEGRA_GPCDMA_CHANNEL_BASE_ADD_OFFSET 0x20000
> >>> +#define TEGRA_GPCDMA_CHANNEL_BASE_ADDR_OFFSET 0x10000
> >>
> >> Why did this value change? There is no mention in the commit message. If
> >> this was incorrect before, then this needs to be a separate patch and
> >> tagged with the appropriate fixes tag so that this can be picked up for
> >> stable.
> > This is mentioned in the commit message.
> >
> > "... and adjusting the register address to skip channel0 ..."
> >
> > Probably it is not very clear that it directs to this change. Shall I update the
> > commit message to have a clearer context?
>
> Ah OK. I was wondering how this worked with 'channel_reg_size' but
> looking closer I see channel_reg_size is always SZ_64K. I wonder why we
> even bother having this parameter and don't use SZ_64K directly?
There is an offset from the base address which the per channel registers start.
Although this offset value happens to match with the channel_reg_size, this is
not actually the per channel register size.
>
> Anyway, for now this is fine.
>
Thanks for the review.

Regards,
Akhil