| Message ID | 20170329160932.27157-1-abailon@baylibre.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
Hello! On 03/29/2017 07:09 PM, Alexandre Bailon wrote: > The CPPI 4.1 DMA is sharing its clock with the USB OTG, > and most of the time, the clock will be enabled by USB. > But during the init of the DMA, USB is not enabled (waiting for DMA), > and then we must enable the clock before to do anything. s/to do/doing/. > Add the clock for the CPPI 4.1 DMA engine. > > Note: > This patch is to apply instead of: > "ARM: davinci: Make the usb20 clock available to PM runtime" > > Signed-off-by: Alexandre Bailon <abailon@baylibre.com> > --- > arch/arm/mach-davinci/da830.c | 1 + > arch/arm/mach-davinci/da850.c | 1 + > arch/arm/mach-davinci/da8xx-dt.c | 2 ++ > 3 files changed, 4 insertions(+) > > diff --git a/arch/arm/mach-davinci/da830.c b/arch/arm/mach-davinci/da830.c > index 073c458..ae4a8a5 100644 > --- a/arch/arm/mach-davinci/da830.c > +++ b/arch/arm/mach-davinci/da830.c > @@ -413,6 +413,7 @@ static struct clk_lookup da830_clks[] = { > CLK("davinci-mcasp.1", NULL, &mcasp1_clk), > CLK("davinci-mcasp.2", NULL, &mcasp2_clk), > CLK("musb-da8xx", "usb20", &usb20_clk), > + CLK("cppi41-dmaengine", NULL, &usb20_clk), > CLK(NULL, "aemif", &aemif_clk), > CLK(NULL, "aintc", &aintc_clk), > CLK(NULL, "secu_mgr", &secu_mgr_clk), > diff --git a/arch/arm/mach-davinci/da850.c b/arch/arm/mach-davinci/da850.c > index ccad2f9..9739aaa 100644 > --- a/arch/arm/mach-davinci/da850.c > +++ b/arch/arm/mach-davinci/da850.c > @@ -560,6 +560,7 @@ static struct clk_lookup da850_clks[] = { > CLK("davinci-nand.0", "aemif", &aemif_nand_clk), > CLK("ohci-da8xx", "usb11", &usb11_clk), > CLK("musb-da8xx", "usb20", &usb20_clk), > + CLK("cppi41-dmaengine", NULL, &usb20_clk), > CLK("spi_davinci.0", NULL, &spi0_clk), > CLK("spi_davinci.1", NULL, &spi1_clk), > CLK("vpif", NULL, &vpif_clk), > diff --git a/arch/arm/mach-davinci/da8xx-dt.c b/arch/arm/mach-davinci/da8xx-dt.c > index e3cef50..5bf08ba 100644 > --- a/arch/arm/mach-davinci/da8xx-dt.c > +++ b/arch/arm/mach-davinci/da8xx-dt.c > @@ -53,6 +53,8 @@ static struct of_dev_auxdata da850_auxdata_lookup[] __initdata = { > OF_DEV_AUXDATA("ti,da830-musb", 0x01e00000, "musb-da8xx", NULL), > OF_DEV_AUXDATA("ti,da830-usb-phy", 0x01c1417c, "da8xx-usb-phy", NULL), > OF_DEV_AUXDATA("ti,da850-ahci", 0x01e18000, "ahci_da850", NULL), > + OF_DEV_AUXDATA("ti,da830-cppi41", 0x01e01000, "cppi41-dma-engine", Not "cppi41-dmaengine"? [...] MBR, Sergei -- To unsubscribe from this list: send the line "unsubscribe linux-omap" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Hello Serge, On 03/29/2017 06:16 PM, Sergei Shtylyov wrote: > Hello! > > On 03/29/2017 07:09 PM, Alexandre Bailon wrote: > >> The CPPI 4.1 DMA is sharing its clock with the USB OTG, >> and most of the time, the clock will be enabled by USB. >> But during the init of the DMA, USB is not enabled (waiting for DMA), >> and then we must enable the clock before to do anything. > > s/to do/doing/. > >> Add the clock for the CPPI 4.1 DMA engine. >> >> Note: >> This patch is to apply instead of: >> "ARM: davinci: Make the usb20 clock available to PM runtime" >> >> Signed-off-by: Alexandre Bailon <abailon@baylibre.com> >> --- >> arch/arm/mach-davinci/da830.c | 1 + >> arch/arm/mach-davinci/da850.c | 1 + >> arch/arm/mach-davinci/da8xx-dt.c | 2 ++ >> 3 files changed, 4 insertions(+) >> >> diff --git a/arch/arm/mach-davinci/da830.c >> b/arch/arm/mach-davinci/da830.c >> index 073c458..ae4a8a5 100644 >> --- a/arch/arm/mach-davinci/da830.c >> +++ b/arch/arm/mach-davinci/da830.c >> @@ -413,6 +413,7 @@ static struct clk_lookup da830_clks[] = { >> CLK("davinci-mcasp.1", NULL, &mcasp1_clk), >> CLK("davinci-mcasp.2", NULL, &mcasp2_clk), >> CLK("musb-da8xx", "usb20", &usb20_clk), >> + CLK("cppi41-dmaengine", NULL, &usb20_clk), >> CLK(NULL, "aemif", &aemif_clk), >> CLK(NULL, "aintc", &aintc_clk), >> CLK(NULL, "secu_mgr", &secu_mgr_clk), >> diff --git a/arch/arm/mach-davinci/da850.c >> b/arch/arm/mach-davinci/da850.c >> index ccad2f9..9739aaa 100644 >> --- a/arch/arm/mach-davinci/da850.c >> +++ b/arch/arm/mach-davinci/da850.c >> @@ -560,6 +560,7 @@ static struct clk_lookup da850_clks[] = { >> CLK("davinci-nand.0", "aemif", &aemif_nand_clk), >> CLK("ohci-da8xx", "usb11", &usb11_clk), >> CLK("musb-da8xx", "usb20", &usb20_clk), >> + CLK("cppi41-dmaengine", NULL, &usb20_clk), >> CLK("spi_davinci.0", NULL, &spi0_clk), >> CLK("spi_davinci.1", NULL, &spi1_clk), >> CLK("vpif", NULL, &vpif_clk), >> diff --git a/arch/arm/mach-davinci/da8xx-dt.c >> b/arch/arm/mach-davinci/da8xx-dt.c >> index e3cef50..5bf08ba 100644 >> --- a/arch/arm/mach-davinci/da8xx-dt.c >> +++ b/arch/arm/mach-davinci/da8xx-dt.c >> @@ -53,6 +53,8 @@ static struct of_dev_auxdata da850_auxdata_lookup[] >> __initdata = { >> OF_DEV_AUXDATA("ti,da830-musb", 0x01e00000, "musb-da8xx", NULL), >> OF_DEV_AUXDATA("ti,da830-usb-phy", 0x01c1417c, "da8xx-usb-phy", >> NULL), >> OF_DEV_AUXDATA("ti,da850-ahci", 0x01e18000, "ahci_da850", NULL), >> + OF_DEV_AUXDATA("ti,da830-cppi41", 0x01e01000, "cppi41-dma-engine", > > Not "cppi41-dmaengine"? Oops. I should I removed this line. It must be made in the da8xx.c USB driver. > > [...] > > MBR, Sergei > Regards, Alexandre -- To unsubscribe from this list: send the line "unsubscribe linux-omap" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Wednesday 29 March 2017 09:39 PM, Alexandre Bailon wrote: > The CPPI 4.1 DMA is sharing its clock with the USB OTG, > and most of the time, the clock will be enabled by USB. > But during the init of the DMA, USB is not enabled (waiting for DMA), > and then we must enable the clock before to do anything. > Add the clock for the CPPI 4.1 DMA engine. > > Note: > This patch is to apply instead of: > "ARM: davinci: Make the usb20 clock available to PM runtime" Okay, but I still liked the fact that that patch was using NULL as con_id for MUSB clock. That makes sense because MUSB gets a single clock input. I think you should still make that change. If not for v4.12, then for v4.13. > > Signed-off-by: Alexandre Bailon <abailon@baylibre.com> > --- > arch/arm/mach-davinci/da830.c | 1 + > arch/arm/mach-davinci/da850.c | 1 + > arch/arm/mach-davinci/da8xx-dt.c | 2 ++ > 3 files changed, 4 insertions(+) > > diff --git a/arch/arm/mach-davinci/da830.c b/arch/arm/mach-davinci/da830.c > index 073c458..ae4a8a5 100644 > --- a/arch/arm/mach-davinci/da830.c > +++ b/arch/arm/mach-davinci/da830.c > @@ -413,6 +413,7 @@ static struct clk_lookup da830_clks[] = { > CLK("davinci-mcasp.1", NULL, &mcasp1_clk), > CLK("davinci-mcasp.2", NULL, &mcasp2_clk), > CLK("musb-da8xx", "usb20", &usb20_clk), > + CLK("cppi41-dmaengine", NULL, &usb20_clk), Did you try reading /sys/kernel/debug/davinci_clocks after this patch? It will hang because of the loop created here. Looks like what you want is cppi4.1 dma clock to be a child of MUSB clock. That way, even if DMA is enabled before MUSB, it still works. Thanks, Sekhar -- To unsubscribe from this list: send the line "unsubscribe linux-omap" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On 04/04/2017 08:16 AM, Sekhar Nori wrote: > On Wednesday 29 March 2017 09:39 PM, Alexandre Bailon wrote: >> The CPPI 4.1 DMA is sharing its clock with the USB OTG, >> and most of the time, the clock will be enabled by USB. >> But during the init of the DMA, USB is not enabled (waiting for DMA), >> and then we must enable the clock before to do anything. >> Add the clock for the CPPI 4.1 DMA engine. >> >> Note: >> This patch is to apply instead of: >> "ARM: davinci: Make the usb20 clock available to PM runtime" > Okay, but I still liked the fact that that patch was using NULL as > con_id for MUSB clock. That makes sense because MUSB gets a single clock > input. I think you should still make that change. If not for v4.12, then > for v4.13. > >> Signed-off-by: Alexandre Bailon <abailon@baylibre.com> >> --- >> arch/arm/mach-davinci/da830.c | 1 + >> arch/arm/mach-davinci/da850.c | 1 + >> arch/arm/mach-davinci/da8xx-dt.c | 2 ++ >> 3 files changed, 4 insertions(+) >> >> diff --git a/arch/arm/mach-davinci/da830.c b/arch/arm/mach-davinci/da830.c >> index 073c458..ae4a8a5 100644 >> --- a/arch/arm/mach-davinci/da830.c >> +++ b/arch/arm/mach-davinci/da830.c >> @@ -413,6 +413,7 @@ static struct clk_lookup da830_clks[] = { >> CLK("davinci-mcasp.1", NULL, &mcasp1_clk), >> CLK("davinci-mcasp.2", NULL, &mcasp2_clk), >> CLK("musb-da8xx", "usb20", &usb20_clk), >> + CLK("cppi41-dmaengine", NULL, &usb20_clk), > Did you try reading /sys/kernel/debug/davinci_clocks after this patch? > It will hang because of the loop created here. I have tried and I did not get any issues. > > Looks like what you want is cppi4.1 dma clock to be a child of MUSB > clock. That way, even if DMA is enabled before MUSB, it still works. OK. I will update the patch to make cppi4.1 clock a child of usb clock. > > Thanks, > Sekhar Thanks, Alexandre -- To unsubscribe from this list: send the line "unsubscribe linux-omap" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/arch/arm/mach-davinci/da830.c b/arch/arm/mach-davinci/da830.c index 073c458..ae4a8a5 100644 --- a/arch/arm/mach-davinci/da830.c +++ b/arch/arm/mach-davinci/da830.c @@ -413,6 +413,7 @@ static struct clk_lookup da830_clks[] = { CLK("davinci-mcasp.1", NULL, &mcasp1_clk), CLK("davinci-mcasp.2", NULL, &mcasp2_clk), CLK("musb-da8xx", "usb20", &usb20_clk), + CLK("cppi41-dmaengine", NULL, &usb20_clk), CLK(NULL, "aemif", &aemif_clk), CLK(NULL, "aintc", &aintc_clk), CLK(NULL, "secu_mgr", &secu_mgr_clk), diff --git a/arch/arm/mach-davinci/da850.c b/arch/arm/mach-davinci/da850.c index ccad2f9..9739aaa 100644 --- a/arch/arm/mach-davinci/da850.c +++ b/arch/arm/mach-davinci/da850.c @@ -560,6 +560,7 @@ static struct clk_lookup da850_clks[] = { CLK("davinci-nand.0", "aemif", &aemif_nand_clk), CLK("ohci-da8xx", "usb11", &usb11_clk), CLK("musb-da8xx", "usb20", &usb20_clk), + CLK("cppi41-dmaengine", NULL, &usb20_clk), CLK("spi_davinci.0", NULL, &spi0_clk), CLK("spi_davinci.1", NULL, &spi1_clk), CLK("vpif", NULL, &vpif_clk), diff --git a/arch/arm/mach-davinci/da8xx-dt.c b/arch/arm/mach-davinci/da8xx-dt.c index e3cef50..5bf08ba 100644 --- a/arch/arm/mach-davinci/da8xx-dt.c +++ b/arch/arm/mach-davinci/da8xx-dt.c @@ -53,6 +53,8 @@ static struct of_dev_auxdata da850_auxdata_lookup[] __initdata = { OF_DEV_AUXDATA("ti,da830-musb", 0x01e00000, "musb-da8xx", NULL), OF_DEV_AUXDATA("ti,da830-usb-phy", 0x01c1417c, "da8xx-usb-phy", NULL), OF_DEV_AUXDATA("ti,da850-ahci", 0x01e18000, "ahci_da850", NULL), + OF_DEV_AUXDATA("ti,da830-cppi41", 0x01e01000, "cppi41-dma-engine", + NULL), {} };
The CPPI 4.1 DMA is sharing its clock with the USB OTG, and most of the time, the clock will be enabled by USB. But during the init of the DMA, USB is not enabled (waiting for DMA), and then we must enable the clock before to do anything. Add the clock for the CPPI 4.1 DMA engine. Note: This patch is to apply instead of: "ARM: davinci: Make the usb20 clock available to PM runtime" Signed-off-by: Alexandre Bailon <abailon@baylibre.com> --- arch/arm/mach-davinci/da830.c | 1 + arch/arm/mach-davinci/da850.c | 1 + arch/arm/mach-davinci/da8xx-dt.c | 2 ++ 3 files changed, 4 insertions(+)