lkml.org 
[lkml]   [2014]   [May]   [5]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
SubjectRe: [PATCH v2 5/6] Documentation: dt: soc: add Keystone Navigator DMA bindings
From
On Wed, Apr 23, 2014 at 6:46 PM, Santosh Shilimkar
<santosh.shilimkar@ti.com> wrote:
> The Keystone Navigator DMA driver sets up the dma channels and flows for
> the QMSS(Queue Manager SubSystem) who triggers the actual data movements
> across clients using destination queues. Every client modules like
> NETCP(Network Coprocessor), SRIO(Serial Rapid IO) and CRYPTO
> Engines has its own instance of packet dma hardware. QMSS has also
> an internal packet DMA module which is used as an infrastructure
> DMA with zero copy.
>
> Initially this driver was proposed as DMA engine driver but since the
> hardware is not typical DMA engine and hence doesn't comply with typical
> DMA engine driver needs, that approach was naked. Link to that
> discussion -
> https://lkml.org/lkml/2014/3/18/340
>
> As aligned, now we pair the Navigator DMA with its companion Navigator
> QMSS subsystem driver.
>
> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> Cc: Kumar Gala <galak@codeaurora.org>
> Cc: Olof Johansson <olof@lixom.net>
> Cc: Arnd Bergmann <arnd@arndb.de>
> Cc: Grant Likely <grant.likely@linaro.org>
> Cc: Rob Herring <robh+dt@kernel.org>
> Cc: Mark Rutland <mark.rutland@arm.com>
> Signed-off-by: Sandeep Nair <sandeep_n@ti.com>
> Signed-off-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
> ---
> .../bindings/soc/keystone-navigator-dma.txt | 101 ++++++++++++++++++++
> 1 file changed, 101 insertions(+)
> create mode 100644 Documentation/devicetree/bindings/soc/keystone-navigator-dma.txt
>
> diff --git a/Documentation/devicetree/bindings/soc/keystone-navigator-dma.txt b/Documentation/devicetree/bindings/soc/keystone-navigator-dma.txt
> new file mode 100644
> index 0000000..4b7d9a7
> --- /dev/null
> +++ b/Documentation/devicetree/bindings/soc/keystone-navigator-dma.txt
> @@ -0,0 +1,101 @@
> +Keystone Navigator DMA Controller
> +
> +This document explains the device tree bindings for the packet dma
> +on keystone devices. The Keystone Navigator DMA driver sets up the dma
> +channels and flows for the QMSS(Queue Manager SubSystem) who triggers
> +the actual data movements across clients using destination queues. Every
> +client modules like NETCP(Network Coprocessor), SRIO(Serial Rapid IO),
> +CRYPTO Engines etc has its own instance of dma hardware. QMSS has also
> +an internal packet DMA module which is used as an infrastructure DMA
> +with zero copy.
> +
> +Navigator DMA cloud layout:
> + ------------------
> + | Navigator DMAs |
> + ------------------
> + |
> + |-> DMA instance #0
> + |
> + |-> DMA instance #1
> + .
> + .
> + |
> + |-> DMA instance #n
> +
> +Navigator DMA properties:
> +Required properties:
> + - compatible: Should be "ti,keystone-navigator-dma"
> + - clocks: phandle to dma instances clocks

2 clocks are in the example. Define how many clocks and the order.

> +
> +DMA instance properties:
> +Required properties:
> + - reg: Should contain register location and length of the following dma
> + register regions. Register regions should be specified in the following
> + order.
> + - Global control register region (global).
> + - Tx DMA channel configuration register region (txchan).
> + - Rx DMA channel configuration register region (rxchan).
> + - Tx DMA channel Scheduler configuration register region (txsched).
> + - Rx DMA flow configuration register region (rxflow).
> + - qm-base-address: Base address of the logical queue managers for dma.

Why is this not part of reg?

> + - #dma-cells: Has to be 1. Keystone DMA doesn't support anything else.

So this uses the common DMA binding?

> +
> +Optional properties:
> + - reg-names: Names for the register regions.
> + - enable-all: Enable all DMA channels.

vs. none of them?

> + - loop-back: To loopback Tx streaming I/F to Rx streaming I/F. Used for
> + infrastructure transfers.
> + - rx-retry-timeout: Number of dma cycles to wait before retry on buffer
> + starvation.

Prefix these with "ti,"

> +Example:
> +
> + knav_dmas: knav_dmas@0 {
> + compatible = "ti,keystone-navigator-dma";
> + clocks = <&papllclk>, <&clkxge>;
> + #address-cells = <1>;
> + #size-cells = <1>;
> + ranges;
> +
> + dma_gbe {
> + dma-id = <0>;

This is not documented in the above binding.

> + reg = <0x2004000 0x100>,
> + <0x2004400 0x120>,
> + <0x2004800 0x300>,
> + <0x2004c00 0x120>,
> + <0x2005000 0x400>;
> + reg-names = "global", "txchan", "rxchan",
> + "txsched", "rxflow";
> +
> + qm-base-address = <0x23a80000 0x23a90000
> + 0x23aa0000 0x23ab0000>;
> + };
> +
> + dma_xgbe {
> + dma-id = <1>;
> + reg = <0x2fa1000 0x100>,
> + <0x2fa1400 0x200>,
> + <0x2fa1800 0x200>,
> + <0x2fa1c00 0x200>,
> + <0x2fa2000 0x400>;
> + reg-names = "global", "txchan", "rxchan",
> + "txsched", "rxflow";
> +
> + qm-base-address = <0x23a80000 0x23a90000
> + 0x23aa0000 0x23ab0000>;

The same address as above is correct here?

> + };
> + };
> +
> +Navigator DMA client:
> +Required properties:
> +- dma-id: DMA instance number

How is this used? It is to match with the above "DMA Instance"? Can
you use phandles?

The use of DMA client and DMA instance are not very clear.

> +- rx-channel: DMA receive channel number
> +- tx-channel: DMA transmit channel number

These could be args to a phandle.

> +
> +Example:
> +
> + netcp: netcp@2090000 {
> + dma-id = <0>;
> + tx-channel = <8>;
> + rx-channel = <22>;
> + ...
> + };
> --
> 1.7.9.5
>


\
 
 \ /
  Last update: 2014-05-05 18:21    [W:0.164 / U:0.032 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site