| Message ID | 20180116190236.14558-1-srinivas.kandagatla@linaro.org |
|---|---|
| Headers | show |
| Series | dmaengine: qcom: bam_dma: fixes for remotely controlled bam | expand |
On 1/16/2018 12:02 PM, srinivas.kandagatla@linaro.org wrote: > From: Srinivas Kandagatla <srinivas.kandagatla@linaro.org> > > When BAM is remotely controlled it does not sound correct to control > its clk on Linux side. Make it optional, so that its not madatory > for remote controlled BAM instances. > > Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org> > --- > drivers/dma/qcom/bam_dma.c | 15 ++++++++------- > 1 file changed, 8 insertions(+), 7 deletions(-) > > diff --git a/drivers/dma/qcom/bam_dma.c b/drivers/dma/qcom/bam_dma.c > index 03c4eb3fd314..78e488e8f96d 100644 > --- a/drivers/dma/qcom/bam_dma.c > +++ b/drivers/dma/qcom/bam_dma.c > @@ -1180,13 +1180,14 @@ static int bam_dma_probe(struct platform_device *pdev) > "qcom,controlled-remotely"); > > bdev->bamclk = devm_clk_get(bdev->dev, "bam_clk"); > - if (IS_ERR(bdev->bamclk)) > - return PTR_ERR(bdev->bamclk); > - > - ret = clk_prepare_enable(bdev->bamclk); > - if (ret) { > - dev_err(bdev->dev, "failed to prepare/enable clock\n"); > - return ret; > + if (IS_ERR(bdev->bamclk)) { > + bdev->bamclk = NULL; > + } else { > + ret = clk_prepare_enable(bdev->bamclk); > + if (ret) { > + dev_err(bdev->dev, "failed to prepare/enable clock\n"); > + return ret; > + } I believe you can also keep pm_runtime disabled if BAM is remotely controlled. Thanks Sagar > } > > ret = bam_init(bdev);
On 16/01/18 19:38, Sagar Dharia wrote: > > On 1/16/2018 12:02 PM, srinivas.kandagatla@linaro.org wrote: >> From: Srinivas Kandagatla <srinivas.kandagatla@linaro.org> >> >> When BAM is remotely controlled it does not sound correct to control >> its clk on Linux side. Make it optional, so that its not madatory >> for remote controlled BAM instances. >> >> Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org> >> --- >> drivers/dma/qcom/bam_dma.c | 15 ++++++++------- >> 1 file changed, 8 insertions(+), 7 deletions(-) >> >> diff --git a/drivers/dma/qcom/bam_dma.c b/drivers/dma/qcom/bam_dma.c >> index 03c4eb3fd314..78e488e8f96d 100644 >> --- a/drivers/dma/qcom/bam_dma.c >> +++ b/drivers/dma/qcom/bam_dma.c >> @@ -1180,13 +1180,14 @@ static int bam_dma_probe(struct >> platform_device *pdev) >> "qcom,controlled-remotely"); >> bdev->bamclk = devm_clk_get(bdev->dev, "bam_clk"); >> - if (IS_ERR(bdev->bamclk)) >> - return PTR_ERR(bdev->bamclk); >> - >> - ret = clk_prepare_enable(bdev->bamclk); >> - if (ret) { >> - dev_err(bdev->dev, "failed to prepare/enable clock\n"); >> - return ret; >> + if (IS_ERR(bdev->bamclk)) { >> + bdev->bamclk = NULL; >> + } else { >> + ret = clk_prepare_enable(bdev->bamclk); >> + if (ret) { >> + dev_err(bdev->dev, "failed to prepare/enable clock\n"); >> + return ret; >> + } > I believe you can also keep pm_runtime disabled if BAM is remotely > controlled. Yes, that's another topic which should be fixed too. I will add that patch in next version. thanks, srini > > Thanks > Sagar >> } >> ret = bam_init(bdev); >
On 17/01/18 10:18, Vinod Koul wrote: > On Tue, Jan 16, 2018 at 07:02:32PM +0000, srinivas.kandagatla@linaro.org wrote: >> From: Srinivas Kandagatla <srinivas.kandagatla@linaro.org> >> >> Hi Andy, >> >> I did hit few issues while trying out SLIMBus BAM on DB820c, this BAM instance >> is remotely controlled and powered up after ADSP is booted using QMI commands. > > What do you mean by "remotely controlled" in this series? DMA controller is controlled by the remote processor, which is a DSP in this case. Linux side is in remote mode in this setup, this can setup transfer descriptors and start/stop a transfer. But all the initialization/powerup part will be done in the remote processor. --srini > >> >> Firstly some of the master registers are written even when the BAM is remotely >> controlled, and secondly reading registers when bam is not ready yet. >> >> These 4 patches address these issues, there are few more issues like doing PM >> in simillar usecase, these will be addressed soon. >> >> Thanks, >> Srini >> >> Srinivas Kandagatla (4): >> dmaengine: qcom: bam_dma: make bam clk optional >> dmaengine: qcom: bam_dma: add num-channels binding for remotely >> controlled >> dmaengine: qcom: bam_dma: do not write to global regs in remote mode >> dmaengine: qcom: bam_dma: Add num-ees dt binding for remotely >> controlled >> >> .../devicetree/bindings/dma/qcom_bam_dma.txt | 4 ++ >> drivers/dma/qcom/bam_dma.c | 56 +++++++++++++++------- >> 2 files changed, 43 insertions(+), 17 deletions(-) >> >> -- >> 2.15.1 >> >
On 19/01/18 05:52, Vinod Koul wrote: > On Tue, Jan 16, 2018 at 07:02:33PM +0000, srinivas.kandagatla@linaro.org wrote: >> From: Srinivas Kandagatla <srinivas.kandagatla@linaro.org> >> >> When BAM is remotely controlled it does not sound correct to control >> its clk on Linux side. Make it optional, so that its not madatory > > s/madatory/mandatory > Yep, >> for remote controlled BAM instances. >> >> Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org> >> --- >> drivers/dma/qcom/bam_dma.c | 15 ++++++++------- >> 1 file changed, 8 insertions(+), 7 deletions(-) >> >> diff --git a/drivers/dma/qcom/bam_dma.c b/drivers/dma/qcom/bam_dma.c >> index 03c4eb3fd314..78e488e8f96d 100644 >> --- a/drivers/dma/qcom/bam_dma.c >> +++ b/drivers/dma/qcom/bam_dma.c >> @@ -1180,13 +1180,14 @@ static int bam_dma_probe(struct platform_device *pdev) >> "qcom,controlled-remotely"); >> >> bdev->bamclk = devm_clk_get(bdev->dev, "bam_clk"); > > but you still do clk_get unconditionally? Only reason to do this way is to not break existing users in the mainline. remotely controlled BAM is already supported in upstream driver, there are users of this who pass clk from device tree, If I make this conditional then subsequent reads to the BAM registers for those instances might crash the system. This sounds wrong to control clk from linux for the dma controller which is remotely controlled. These users should be transitioned to new bindings once the new bindings endup in the mainline. > >> - if (IS_ERR(bdev->bamclk)) >> - return PTR_ERR(bdev->bamclk); >> - >> - ret = clk_prepare_enable(bdev->bamclk); >> - if (ret) { >> - dev_err(bdev->dev, "failed to prepare/enable clock\n"); >> - return ret; >> + if (IS_ERR(bdev->bamclk)) { >> + bdev->bamclk = NULL; >> + } else { >> + ret = clk_prepare_enable(bdev->bamclk); >> + if (ret) { >> + dev_err(bdev->dev, "failed to prepare/enable clock\n"); >> + return ret; >> + } > > wouldn't it be better to set that an instance is remote controlled and thus > not at all visible to Linux? We already have a flag "controlled_remotely" for that in the driver. thanks, srini > >> } >> >> ret = bam_init(bdev); >> -- >> 2.15.1 >> >> -- >> To unsubscribe from this list: send the line "unsubscribe dmaengine" in >> the body of a message to majordomo@vger.kernel.org >> More majordomo info at http://vger.kernel.org/majordomo-info.html >
On 19/01/18 05:55, Vinod Koul wrote: > On Tue, Jan 16, 2018 at 07:02:34PM +0000, srinivas.kandagatla@linaro.org wrote: >> From: Srinivas Kandagatla <srinivas.kandagatla@linaro.org> >> >> When Linux is master of BAM, it can directly read registers to know number >> of supported channels, however when its remotely controlled reading these >> registers would trigger a crash if the BAM is not yet intialized/powered up >> on the remote side. >> >> This patch adds num-channels binding to specify number of supported >> dma channels on remotely controlled BAM. >> >> Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org> >> --- >> Documentation/devicetree/bindings/dma/qcom_bam_dma.txt | 2 ++ >> drivers/dma/qcom/bam_dma.c | 13 +++++++++++-- >> 2 files changed, 13 insertions(+), 2 deletions(-) >> >> diff --git a/Documentation/devicetree/bindings/dma/qcom_bam_dma.txt b/Documentation/devicetree/bindings/dma/qcom_bam_dma.txt >> index 9cbf5d9df8fd..aa6822cbb230 100644 >> --- a/Documentation/devicetree/bindings/dma/qcom_bam_dma.txt >> +++ b/Documentation/devicetree/bindings/dma/qcom_bam_dma.txt >> @@ -15,6 +15,8 @@ Required properties: >> the secure world. >> - qcom,controlled-remotely : optional, indicates that the bam is controlled by >> remote proccessor i.e. execution environment. >> +- num-channels : optional, indicates supported number of DMA channels in a >> + remotely controlled bam. >> >> Example: >> >> diff --git a/drivers/dma/qcom/bam_dma.c b/drivers/dma/qcom/bam_dma.c >> index 78e488e8f96d..523bd178047a 100644 >> --- a/drivers/dma/qcom/bam_dma.c >> +++ b/drivers/dma/qcom/bam_dma.c >> @@ -1083,8 +1083,10 @@ static int bam_init(struct bam_device *bdev) >> if (bdev->ee >= val) >> return -EINVAL; >> >> - val = readl_relaxed(bam_addr(bdev, 0, BAM_NUM_PIPES)); >> - bdev->num_channels = val & BAM_NUM_PIPES_MASK; >> + if (!bdev->num_channels) { >> + val = readl_relaxed(bam_addr(bdev, 0, BAM_NUM_PIPES)); >> + bdev->num_channels = val & BAM_NUM_PIPES_MASK; >> + } >> >> if (bdev->controlled_remotely) >> return 0; >> @@ -1179,6 +1181,13 @@ static int bam_dma_probe(struct platform_device *pdev) >> bdev->controlled_remotely = of_property_read_bool(pdev->dev.of_node, >> "qcom,controlled-remotely"); >> >> + if (bdev->controlled_remotely) { > > hmm so if we remove the remotely controlled instanced from DT and then Linux > won't see them and not do anything. Do we need to do configuration of these > instances too? No, bindings "num-channels" and "num-ees" are applicable to remote controlled instances only, which is documented in the DT bindings. Normally these values come from register reads, which are not possible in case of remotely controlled instances as BAM driver would not know if the remote is powered up or not. thanks, srini > >> + ret = of_property_read_u32(pdev->dev.of_node, "num-channels", >> + &bdev->num_channels); >> + if (ret) >> + dev_err(bdev->dev, "num-channels unspecified in dt\n"); >> + } >> + >> bdev->bamclk = devm_clk_get(bdev->dev, "bam_clk"); >> if (IS_ERR(bdev->bamclk)) { >> bdev->bamclk = NULL; >> -- >> 2.15.1 >> >
From: Srinivas Kandagatla <srinivas.kandagatla@linaro.org> Hi Andy, I did hit few issues while trying out SLIMBus BAM on DB820c, this BAM instance is remotely controlled and powered up after ADSP is booted using QMI commands. Firstly some of the master registers are written even when the BAM is remotely controlled, and secondly reading registers when bam is not ready yet. These 4 patches address these issues, there are few more issues like doing PM in simillar usecase, these will be addressed soon. Thanks, Srini Srinivas Kandagatla (4): dmaengine: qcom: bam_dma: make bam clk optional dmaengine: qcom: bam_dma: add num-channels binding for remotely controlled dmaengine: qcom: bam_dma: do not write to global regs in remote mode dmaengine: qcom: bam_dma: Add num-ees dt binding for remotely controlled .../devicetree/bindings/dma/qcom_bam_dma.txt | 4 ++ drivers/dma/qcom/bam_dma.c | 56 +++++++++++++++------- 2 files changed, 43 insertions(+), 17 deletions(-) -- 2.15.1