> On 07/07/2022 11:27, Akhil R wrote:
> > In certain cases where the DMA client bus gets corrupted or if the end
> > device ceases to send/receive data, DMA can wait indefinitely for the
> > data to be received/sent. Attempting to terminate the transfer will
> > put the DMA in pause flush mode and it remains there.
> >
> > The channel is irrecoverable once this pause times out in Tegra194 and
> > earlier chips. Whereas, from Tegra234, it can be recovered by
> > disabling the channel and reprograming it.
> >
> > Hence add a new terminate() function that ignores the outcome of
> > dma_pause() and disables the channel.
> >
> > Signed-off-by: Akhil R <akhilrajeev@xxxxxxxxxx>
> > ---
> > arch/arm64/boot/dts/nvidia/tegra234.dtsi | 5 +++--
> > drivers/dma/tegra186-gpc-dma.c | 26 ++++++++++++++++++++++--
> > 2 files changed, 27 insertions(+), 4 deletions(-)
> >
> > diff --git a/arch/arm64/boot/dts/nvidia/tegra234.dtsi
> > b/arch/arm64/boot/dts/nvidia/tegra234.dtsi
> > index cf611eff7f6b..83d1ad7d3c8c 100644
> > --- a/arch/arm64/boot/dts/nvidia/tegra234.dtsi
> > +++ b/arch/arm64/boot/dts/nvidia/tegra234.dtsi
> > @@ -22,8 +22,9 @@
> > ranges = <0x0 0x0 0x0 0x40000000>;
> >
> > gpcdma: dma-controller@2600000 {
> > - compatible = "nvidia,tegra194-gpcdma",
> > - "nvidia,tegra186-gpcdma";
> > + compatible = "nvidia,tegra234-gpcdma",
> > + "nvidia,tegra194-gpcdma",
> > + "nvidia,tegra186-gpcdma";
> > reg = <0x2600000 0x210000>;
> > resets = <&bpmp TEGRA234_RESET_GPCDMA>;
> > reset-names = "gpcdma";
>
> I think that this should be split into two patches.
>
> > diff --git a/drivers/dma/tegra186-gpc-dma.c
> > b/drivers/dma/tegra186-gpc-dma.c index 05cd451f541d..fa9bda4a2bc6
> > 100644
> > --- a/drivers/dma/tegra186-gpc-dma.c
> > +++ b/drivers/dma/tegra186-gpc-dma.c
> > @@ -157,8 +157,8 @@
> > * If any burst is in flight and DMA paused then this is the time to complete
> > * on-flight burst and update DMA status register.
> > */
> > -#define TEGRA_GPCDMA_BURST_COMPLETE_TIME 20
> > -#define TEGRA_GPCDMA_BURST_COMPLETION_TIMEOUT 100
> > +#define TEGRA_GPCDMA_BURST_COMPLETE_TIME 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
> > @@ -432,6 +432,17 @@ static int tegra_dma_device_resume(struct
> dma_chan *dc)
> > return 0;
> > }
> >
> > +static inline int tegra_dma_pause_noerr(struct tegra_dma_channel
> > +*tdc) {
> > + /* Return 0 irrespective of PAUSE status.
> > + * This is useful to recover channels that can exit out of flush
> > + * state when the channel is disabled.
> > + */
> > +
> > + tegra_dma_pause(tdc);
> > + return 0;
> > +}
>
> The commit message says that "add a new terminate() function that ignores the
> outcome of dma_pause() and disables the channel". But I only see pause being
> done here.
The function is set as .terminate() function in chip_data and is called during
terminate_all(). Since this return 0, tegra_dma_terminate_all() will proceed
and calls tegra_dma_disable() in the next step.
Regards,
Akhil
--
nvpublic
