ARM: dts: imx: Remove custom dma-channel bindings
PATCH 1/2 ARM: dts: imx: Remove custom dma-channel bindings
Commit f30fb03 (ARM: dts: add generic DMA device tree binding for mxs-dma) added the generic DMA dt bindings, but missed to remove the old ones.
Remove the the old dma bindings.
--- arch/arm/boot/dts/imx23.dtsi | 3 --- arch/arm/boot/dts/imx28.dtsi | 5 ----- arch/arm/boot/dts/imx6qdl.dtsi | 1 - 3 files changed, 9 deletions(-)
arch/arm/boot/dts/imx23.dtsi +0 -3
@@ -78,7 +78,6 @@ clock-names = "gpmi_io"; dmas = <&dma_apbh 4>; dma-names = "rx-tx"; - fsl,gpmi-dma-channel = <4>; status = "disabled"; };
@@ -88,7 +87,6 @@ clocks = <&clks 33>; dmas = <&dma_apbh 1>; dma-names = "rx-tx"; - fsl,ssp-dma-channel = <1>; status = "disabled"; };
@@ -366,7 +364,6 @@ clocks = <&clks 33>; dmas = <&dma_apbh 2>; dma-names = "rx-tx"; - fsl,ssp-dma-channel = <2>; status = "disabled"; };
arch/arm/boot/dts/imx28.dtsi +0 -5
@@ -100,7 +100,6 @@ clock-names = "gpmi_io"; dmas = <&dma_apbh 4>; dma-names = "rx-tx"; - fsl,gpmi-dma-channel = <4>; status = "disabled"; };
@@ -112,7 +111,6 @@ clocks = <&clks 46>; dmas = <&dma_apbh 0>; dma-names = "rx-tx"; - fsl,ssp-dma-channel = <0>; status = "disabled"; };
@@ -124,7 +122,6 @@ clocks = <&clks 47>; dmas = <&dma_apbh 1>; dma-names = "rx-tx"; - fsl,ssp-dma-channel = <1>; status = "disabled"; };
@@ -136,7 +133,6 @@ clocks = <&clks 48>; dmas = <&dma_apbh 2>; dma-names = "rx-tx"; - fsl,ssp-dma-channel = <2>; status = "disabled"; };
@@ -148,7 +144,6 @@ clocks = <&clks 49>; dmas = <&dma_apbh 3>; dma-names = "rx-tx"; - fsl,ssp-dma-channel = <3>; status = "disabled"; };
arch/arm/boot/dts/imx6qdl.dtsi +0 -1
@@ -89,7 +89,6 @@ "gpmi_bch_apb", "per1_bch"; dmas = <&dma_apbh 0>; dma-names = "rx-tx"; - fsl,gpmi-dma-channel = <0>; status = "disabled"; };
Signature · 1 line
1.8.1.2
PATCH 2/2 ARM: dts: imx28: Adjust i2c interrupt bindings
Commit e5aba13 (i2c: i2c-mxs: move to use generic DMA helper) changed i2c-mxs.txt and drivers/i2c/busses/i2c-mxs.c to only use one I2C interrupt.
Adapt the dtsi bindings with such change.
--- arch/arm/boot/dts/imx28.dtsi | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
arch/arm/boot/dts/imx28.dtsi +2 -2
@@ -878,7 +878,7 @@ #size-cells = <0>; compatible = "fsl,imx28-i2c"; reg = <0x80058000 0x2000>; - interrupts = <111 68>; + interrupts = <111>; clock-frequency = <100000>; dmas = <&dma_apbx 6>; dma-names = "rx-tx";
@@ -891,7 +891,7 @@ #size-cells = <0>; compatible = "fsl,imx28-i2c"; reg = <0x8005a000 0x2000>; - interrupts = <110 69>; + interrupts = <110>; clock-frequency = <100000>; dmas = <&dma_apbx 7>; dma-names = "rx-tx";
Signature · 1 line
1.8.1.2
Fabio Estevam
reviewerPATCH 2/2 ARM: dts: imx28: Adjust i2c interrupt bindings
Quoted 24 lines · Fabio Estevam
Commit e5aba13 (i2c: i2c-mxs: move to use generic DMA helper) changed i2c-mxs.txt and
drivers/i2c/busses/i2c-mxs.cto only use one I2C interrupt.Adapt the dtsi bindings with such change.
Signed-off-by Fabio Estevam <[email protected]>--- arch/arm/boot/dts/imx28.dtsi | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)arch/arm/boot/dts/imx28.dtsi +1
-1@@ -878,7 +878,7 @@ #size-cells = <0>; compatible = "fsl,imx28-i2c"; reg = <0x80058000 0x2000>; - interrupts = <111 68>; + interrupts = <111>; clock-frequency = <100000>; dmas = <&dma_apbx 6>; dma-names = "rx-tx";
arch/arm/boot/dts/imx28.dtsi line 897
Quoted 8 lines
arch/arm/boot/dts/imx28.dtsi +1
-1@@ -891,7 +891,7 @@ #size-cells = <0>; compatible = "fsl,imx28-i2c"; reg = <0x8005a000 0x2000>; - interrupts = <110 69>; + interrupts = <110>; clock-frequency = <100000>; dmas = <&dma_apbx 7>; dma-names = "rx-tx";
Actually the correct is:
arch/arm/boot/dts/imx28.dtsi +2 -2
@@ -878,7 +878,7 @@ #size-cells = <0>; compatible = "fsl,imx28-i2c"; reg = <0x80058000 0x2000>; - interrupts = <111 68>; + interrupts = <68>; clock-frequency = <100000>; dmas = <&dma_apbx 6>; dma-names = "rx-tx";
@@ -891,7 +891,7 @@ #size-cells = <0>; compatible = "fsl,imx28-i2c"; reg = <0x8005a000 0x2000>; - interrupts = <110 69>; + interrupts = <69>; clock-frequency = <100000>; dmas = <&dma_apbx 7>; dma-names = "rx-tx";
,but then i2c fails to probe.
Has anyone tried mxs i2c dma recently?
Here is what I do on linux-next-20130516:
- Revert b871f1ad3c8a1ac2 (will send this later to alsa-devel)
- Booting mx28evk will show:
[ 6.378298] mmcblk0: p1 p2 p3 [ 29.547791] sgtl5000 0-000a: Failed to get supply 'VDDD': -19 [ 29.556308] 0-000a: 1200 mV normal [ 29.562225] sgtl5000 0-000a: Using internal LDO instead of VDDD [ 29.573156] sgtl5000 0-000a: sgtl5000 revision 0x11
,ie, i2c is too slow as it takes 13 seconds to read the sgtl5000 codec registers. Looks like i2c dma is disabled
Then I apply
arch/arm/boot/dts/imx28.dtsi +2 -2
@@ -878,7 +878,7 @@ #size-cells = <0>; compatible = "fsl,imx28-i2c"; reg = <0x80058000 0x2000>; - interrupts = <111 68>; + interrupts = <68>; clock-frequency = <100000>; dmas = <&dma_apbx 6>; dma-names = "rx-tx";
@@ -891,7 +891,7 @@ #size-cells = <0>; compatible = "fsl,imx28-i2c"; reg = <0x8005a000 0x2000>; - interrupts = <110 69>; + interrupts = <69>; clock-frequency = <100000>; dmas = <&dma_apbx 7>; dma-names = "rx-tx";
and the sgtl5000 probe fails.
Fabio Estevam
reviewerPATCH 2/2 ARM: dts: imx28: Adjust i2c interrupt bindings
Quoted 14 lines · Fabio Estevam
Actually the correct is:
arch/arm/boot/dts/imx28.dtsi +1
-1@@ -878,7 +878,7 @@ #size-cells = <0>; compatible = "fsl,imx28-i2c"; reg = <0x80058000 0x2000>; - interrupts = <111 68>; + interrupts = <68>; clock-frequency = <100000>; dmas = <&dma_apbx 6>; dma-names = "rx-tx";
arch/arm/boot/dts/imx28.dtsi line 897
Quoted 8 lines
arch/arm/boot/dts/imx28.dtsi +1
-1@@ -891,7 +891,7 @@ #size-cells = <0>; compatible = "fsl,imx28-i2c"; reg = <0x8005a000 0x2000>; - interrupts = <110 69>; + interrupts = <69>; clock-frequency = <100000>; dmas = <&dma_apbx 7>; dma-names = "rx-tx";
Sorry, the original patch is correct.
i2c slowness is a different issue.
PATCH 2/2 ARM: dts: imx28: Adjust i2c interrupt bindings
Hi Fabio,
Quoted 3 lines · Fabio Estevam
Sorry, the original patch is correct.
i2c slowness is a different issue.
Would that be related to the timeouts I'm observing ?
http://www.spinics.net/lists/linux-i2c/msg11985.html
I didn't have the time to have a look at it yet. I was planning to hook up a logic analyzer someday in june to debug it. It seems that there is a 1 second timeout with each i2c message that are sent.
Don't hesitate to contact me if you need more info.
Regards,
Signature · 3 lines
Alexandre Belloni, Free Electrons Embedded Linux, Kernel and Android engineering http://free-electrons.com
Fabio Estevam
reviewerPATCH 2/2 ARM: dts: imx28: Adjust i2c interrupt bindings
Hi Alexandre,
Quoted 7 lines · Alexandre Belloni
Would that be related to the timeouts I'm observing ?
http://www.spinics.net/lists/linux-i2c/msg11985.html
I didn't have the time to have a look at it yet. I was planning to hook up a logic analyzer someday in june to debug it. It seems that there is a 1 second timeout with each i2c message that are sent.
Yes, I see the same one second timeout for each i2c transaction here.
Regards,
Fabio Estevam
PATCH 2/2 ARM: dts: imx28: Adjust i2c interrupt bindings
In reply to Fabio EstevamHi Fabio,
Quoted 1 line
i2c slowness is a different issue.
Same happens here for my i.M23 based platform. It seems the PIO mode does not work, or at least not like it works on a i.MX28. Each short transfer needs about one second (without an error message) but does not send anything on the I2C lines.
I need the following patches to make I2C master work within a 3.10-rc2 kernel:
Subject: [PATCH] I2C/MXS: distinguish i.MX23 and i.MX28 based I2C controller
--- drivers/i2c/busses/i2c-mxs.c | 39 +++++++++++++++++++++++++++++++++------ 1 file changed, 33 insertions(+), 6 deletions(-)
drivers/i2c/busses/i2c-mxs.c +33 -6
@@ -97,10 +97,17 @@#define MXS_CMD_I2C_READ (MXS_I2C_CTRL0_SEND_NAK_ON_LAST | \ MXS_I2C_CTRL0_MASTER_MODE) +enum mxs_i2c_devtype { +MXS_I2C_UNKNOWN =0 , +MXS_I2C_V1 , +MXS_I2C_V2 , +}; +/** * struct mxs_i2c_dev - per device, private MXS-I2C data * * @dev: driver model device node +* @dev_type: distinguish i.MX23/i.MX28 features * @regs: IO registers pointer * @cmd_complete: completion object for transaction wait * @cmd_err: error code for last transaction
@@ -108,6 +115,7 @@*/ struct mxs_i2c_dev {struct device *dev; +enum mxs_i2c_devtype dev_type;void __iomem *regs;struct completion cmd_complete;int cmd_err;
@@ -633,8 +641,28 @@ static int mxs_i2c_get_ofdata(struct mxs_i2c_dev *i2c)return 0 ; } +static struct platform_device_id mxs_i2c_devtype[] = { + { + .name ="imx23-i2c" , + .driver_data =MXS_I2C_V1 , + }, { + .name ="imx28-i2c" , + .driver_data =MXS_I2C_V2 , + }, {/* sentinel */ } +}; +MODULE_DEVICE_TABLE (platform, mxs_i2c_devtype); + +static const struct of_device_id mxs_i2c_dt_ids[] = { + { .compatible ="fsl,imx23-i2c" , .data = &mxs_i2c_devtype[0 ], }, + { .compatible ="fsl,imx28-i2c" , .data = &mxs_i2c_devtype[1 ], }, + {/* sentinel */ } +}; +MODULE_DEVICE_TABLE (of, mxs_i2c_dt_ids); +static int mxs_i2c_probe (struct platform_device *pdev) { +const struct of_device_id *of_id = +of_match_device (mxs_i2c_dt_ids, &pdev->dev);struct device *dev = &pdev->dev;struct mxs_i2c_dev *i2c;struct i2c_adapter *adap;
@@ -651,6 +679,11 @@ static int mxs_i2c_probe(struct platform_device *pdev)if (!i2c)return -ENOMEM ; +if (of_id) { +const struct platform_device_id *device_id = of_id->data; + i2c->dev_type = device_id->driver_data; + } + res =platform_get_resource (pdev,IORESOURCE_MEM ,0 ); irq =platform_get_irq (pdev,0 );
@@ -726,12 +759,6 @@ static int mxs_i2c_remove(struct platform_device *pdev)return 0 ; } -static const struct of_device_id mxs_i2c_dt_ids[] = { - { .compatible ="fsl,imx28-i2c" , }, - {/* sentinel */ } -}; -MODULE_DEVICE_TABLE (of, mxs_i2c_dt_ids); -static struct platform_driver mxs_i2c_driver = { .driver = { .name =DRIVER_NAME ,
---------------------------------------------------------------------------------------------
Subject: [PATCH] I2C/MXS: on i.MX23 only do DMA based I2C transfers
--- drivers/i2c/busses/i2c-mxs.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/i2c/busses/i2c-mxs.c +2 -1
@@ -500,9 +500,10 @@ static int mxs_i2c_xfer_msg(struct i2c_adapter *adap, struct i2c_msg *msg,* is set to 8 bytes, transfers shorter than 8 bytes are transfered * using PIO mode while longer transfers use DMA. The 8 byte border is * based on this empirical measurement and a lot of previous frobbing. +* Note: this special feature only works on i.MX28 SoC */ i2c->cmd_err =0 ; -if (msg->len <8 ) { +if (i2c->dev_type ==MXS_I2C_V2 && msg->len <8 ) { ret =mxs_i2c_pio_setup_xfer (adap, msg, flags);if (ret)mxs_i2c_reset (i2c);
Regards,
Juergen
Signature · 4 lines
Pengutronix e.K. | Juergen Beisert | Linux Solutions for Science and Industry | Phone: +49-5121-206917-5128 | Peiner Str. 6-8, 31137 Hildesheim, Germany | Fax: +49-5121-206917-5555 | Amtsgericht Hildesheim, HRA 2686 | http://www.pengutronix.de/ |
PATCH 2/2 ARM: dts: imx28: Adjust i2c interrupt bindings
Dear Juergen Beisert,
Quoted 9 lines
Hi Fabio,
Quoted 1 line
i2c slowness is a different issue.
Same happens here for my i.M23 based platform. It seems the PIO mode does not work, or at least not like it works on a i.MX28. Each short transfer needs about one second (without an error message) but does not send anything on the I2C lines.
Quoted 4 lines
I need the following patches to make I2C master work within a 3.10-rc2 kernel:
Subject: [PATCH] I2C/MXS: distinguish i.MX23 and i.MX28 based I2C
I'm all for it, but then ... won't it be better if you actually fixed the PIO and mixed-mode on MX23 instead of implementing such hack? It might actually turn out to be beneficial for MX28 too.
Best regards,
Marek Vasut
PATCH 2/2 ARM: dts: imx28: Adjust i2c interrupt bindings
Hi Marek,
Quoted 7 lines · Marek Vasut
Quoted 7 lines
Quoted 1 line
i2c slowness is a different issue.
Same happens here for my i.M23 based platform. It seems the PIO mode does not work, or at least not like it works on a i.MX28. Each short transfer needs about one second (without an error message) but does not send anything on the I2C lines.
Quoted 7 lines
Quoted 4 lines
I need the following patches to make I2C master work within a 3.10-rc2 kernel:
Subject: [PATCH] I2C/MXS: distinguish i.MX23 and i.MX28 based I2C
I'm all for it, but then ... won't it be better if you actually fixed the PIO and mixed-mode on MX23 instead of implementing such hack?
If the PIO mode or my patch is a hack depends on the point of view: Lucas told me the PIO mode is *mentioned* but *not specified* in the i.MX23/i.MX28 datasheets.
So, the PIO mode seems to depend on some undocumented status bits in the i.MX28 I2C controller implementation.
Regards,
Juergen
Signature · 4 lines
Pengutronix e.K. | Juergen Beisert | Linux Solutions for Science and Industry | Phone: +49-5121-206917-5128 | Peiner Str. 6-8, 31137 Hildesheim, Germany | Fax: +49-5121-206917-5555 | Amtsgericht Hildesheim, HRA 2686 | http://www.pengutronix.de/ |
PATCH 2/2 ARM: dts: imx28: Adjust i2c interrupt bindings
Dear Juergen Beisert,
Quoted 13 lines
Hi Marek,
Quoted 10 lines · Marek Vasut
Quoted 10 lines
Quoted 1 line
i2c slowness is a different issue.
Same happens here for my i.M23 based platform. It seems the PIO mode does not work, or at least not like it works on a i.MX28. Each short transfer needs about one second (without an error message) but does not send anything on the I2C lines.
I need the following patches to make I2C master work within a 3.10-rc2 kernel:
Quoted 8 lines
Quoted 4 lines
Quoted 1 line
Subject: [PATCH] I2C/MXS: distinguish i.MX23 and i.MX28 based I2C
I'm all for it, but then ... won't it be better if you actually fixed the PIO and mixed-mode on MX23 instead of implementing such hack?
If the PIO mode or my patch is a hack depends on the point of view: Lucas told me the PIO mode is *mentioned* but *not specified* in the i.MX23/i.MX28 datasheets.
The PIO works the same way DMA does -- set up bits and then pump data into the DATA register.
Quoted 2 lines
So, the PIO mode seems to depend on some undocumented status bits in the i.MX28 I2C controller implementation.
How would DMA work then if it used undocumented registers ? It's in the documentation, just read it or ask FSL ;-)
Best regards,
Marek Vasut
PATCH 2/2 ARM: dts: imx28: Adjust i2c interrupt bindings
Am Donnerstag, den 23.05.2013, 16:48 +0200 schrieb Marek Vasut:
Quoted 24 lines
Dear Juergen Beisert,
Quoted 13 lines
Hi Marek,
Quoted 10 lines · Marek Vasut
Quoted 10 lines
Quoted 1 line
i2c slowness is a different issue.
Same happens here for my i.M23 based platform. It seems the PIO mode does not work, or at least not like it works on a i.MX28. Each short transfer needs about one second (without an error message) but does not send anything on the I2C lines.
I need the following patches to make I2C master work within a 3.10-rc2 kernel:
Quoted 8 lines
Quoted 4 lines
Quoted 1 line
Subject: [PATCH] I2C/MXS: distinguish i.MX23 and i.MX28 based I2C
I'm all for it, but then ... won't it be better if you actually fixed the PIO and mixed-mode on MX23 instead of implementing such hack?
If the PIO mode or my patch is a hack depends on the point of view: Lucas told me the PIO mode is *mentioned* but *not specified* in the i.MX23/i.MX28 datasheets.
Quoted 9 lines
The PIO works the same way DMA does -- set up bits and then pump data into the DATA register.
Quoted 2 lines
So, the PIO mode seems to depend on some undocumented status bits in the i.MX28 I2C controller implementation.
How would DMA work then if it used undocumented registers ? It's in the documentation, just read it or ask FSL ;-)
While the PIO mode might use the same controller mechanisms as the DMA mode, PIO mode is _not_ a documented mode of operation for the i.MX23.
To quote the i.MX28 RM: "The I2C block on the i.MX28 supports a new PIO mode or soft-DMA mode.", which implies the PIO mode to be a new mode of operation not found on earlier i.MX SoCs.
The doc is slightly fuzzy here as the i.MX23 RM in contrary states:
"Short transmission (up to three bytes plus address) can be easily triggered using only PIO operations, i.e., no DMA setup required." But again it's not a documented mode of operation, i.MX23 doc only describes the DMA mode.
So while we _might_ be able to get the PIO mode to work on the i.MX23 there is nothing in the doc stating that it's even meant to work. Even while PIO and DMA mode use the same internal mechanisms, there's still plenty of opportunities of fail in there. After all PIO mode relies on reading a debug register in the course of normal operation.
Only more extensive experimentation could show if we are in fact able to make it work, a first shot of using PIO mode on MX23 failed, so it might as well be that Juergens quick fix is correct and we have to disable PIO mode on MX23 altogether. That said please stop slapping the word "hack" over this patch until proven otherwise.
Regards,
Lucas
Signature · 4 lines
Pengutronix e.K. | Lucas Stach | Industrial Linux Solutions | http://www.pengutronix.de/ | Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-5076 | Amtsgericht Hildesheim, HRA 2686 | Fax: +49-5121-206917-5555 |
PATCH 2/2 ARM: dts: imx28: Adjust i2c interrupt bindings
Hi,
Quoted 46 lines · Lucas Stach
Am Donnerstag, den 23.05.2013, 16:48 +0200 schrieb Marek Vasut:
Quoted 20 lines
Dear Juergen Beisert,
Quoted 12 lines
Hi Marek,
Quoted 9 lines · Marek Vasut
Quoted 9 lines
Quoted 1 line
i2c slowness is a different issue.
Same happens here for my i.M23 based platform. It seems the PIO mode does not work, or at least not like it works on a i.MX28. Each short transfer needs about one second (without an error message) but does not send anything on the I2C lines.
I need the following patches to make I2C master work within a 3.10-rc2 kernel:
Quoted 6 lines
Quoted 3 lines
Quoted 1 line
Subject: [PATCH] I2C/MXS: distinguish i.MX23 and i.MX28 based I2C
I'm all for it, but then ... won't it be better if you actually fixed the PIO and mixed-mode on MX23 instead of implementing such hack?
If the PIO mode or my patch is a hack depends on the point of view: Lucas told me the PIO mode is *mentioned* but *not specified* in the i.MX23/i.MX28 datasheets.
Quoted 7 lines
The PIO works the same way DMA does -- set up bits and then pump data into the DATA register.
Quoted 2 lines
So, the PIO mode seems to depend on some undocumented status bits in the i.MX28 I2C controller implementation.
How would DMA work then if it used undocumented registers ? It's in the documentation, just read it or ask FSL ;-)
While the PIO mode might use the same controller mechanisms as the DMA mode, PIO mode is _not_ a documented mode of operation for the i.MX23.
To quote the i.MX28 RM: "The I2C block on the i.MX28 supports a new PIO mode or soft-DMA mode.", which implies the PIO mode to be a new mode of operation not found on earlier i.MX SoCs.
The doc is slightly fuzzy here as the i.MX23 RM in contrary states:
"Short transmission (up to three bytes plus address) can be easily triggered using only PIO operations, i.e., no DMA setup required." But again it's not a documented mode of operation, i.MX23 doc only describes the DMA mode.So while we _might_ be able to get the PIO mode to work on the i.MX23 there is nothing in the doc stating that it's even meant to work. Even while PIO and DMA mode use the same internal mechanisms, there's still plenty of opportunities of fail in there. After all PIO mode relies on reading a debug register in the course of normal operation.
Quoted 5 lines
Only more extensive experimentation could show if we are in fact able to make it work, a first shot of using PIO mode on MX23 failed, so it might as well be that Juergens quick fix is correct and we have to disable PIO mode on MX23 altogether. That said please stop slapping the word "hack" over this patch until proven otherwise.
Still, Fabio and I have been experiencing the one second delay on i.mx28. From my experience, it appeared between 3.7 and 3.9, see:
http://www.spinics.net/lists/linux-i2c/msg11985.html. I still didn't take the time to investigate more yet but it confirms that PIO mode is working on i.mx28 though with an annoying delay.
Regards,
Signature · 3 lines
Alexandre Belloni, Free Electrons Embedded Linux, Kernel and Android engineering http://free-electrons.com
PATCH 2/2 ARM: dts: imx28: Adjust i2c interrupt bindings
Am Donnerstag, den 23.05.2013, 19:51 +0200 schrieb Alexandre Belloni:
Quoted 2 lines
Hi,
[...]
Quoted 13 lines
Quoted 7 lines
So while we _might_ be able to get the PIO mode to work on the i.MX23 there is nothing in the doc stating that it's even meant to work. Even while PIO and DMA mode use the same internal mechanisms, there's still plenty of opportunities of fail in there. After all PIO mode relies on reading a debug register in the course of normal operation.
Quoted 5 lines
Only more extensive experimentation could show if we are in fact able to make it work, a first shot of using PIO mode on MX23 failed, so it might as well be that Juergens quick fix is correct and we have to disable PIO mode on MX23 altogether. That said please stop slapping the word "hack" over this patch until proven otherwise.
Quoted 6 lines
Still, Fabio and I have been experiencing the one second delay on i.mx28. From my experience, it appeared between 3.7 and 3.9, see:
http://www.spinics.net/lists/linux-i2c/msg11985.html. I still didn't take the time to investigate more yet but it confirms that PIO mode is working on i.mx28 though with an annoying delay.
Just for the record: I know that PIO mode is functional on i.MX28 because I have it working on one of our customer boards. I am rather much surprised it's still not working well for you, but haven't had a chance to investigate yet. The i.MX28 RM explicitly describes how to work with the PIO mode, so if it's not working this is bug somewhere.
This thread/patch is about a whole different problem, namely PIO mode being non-functional on i.MX23. As the i.MX23 RM doesn't describe the PIO mode, there is no indication that it's even meant to work on this chip.
Regards,
Lucas
Signature · 4 lines
Pengutronix e.K. | Lucas Stach | Industrial Linux Solutions | http://www.pengutronix.de/ | Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-5076 | Amtsgericht Hildesheim, HRA 2686 | Fax: +49-5121-206917-5555 |
PATCH 2/2 ARM: dts: imx28: Adjust i2c interrupt bindings
Dear Lucas Stach,
Quoted 23 lines
Am Donnerstag, den 23.05.2013, 19:51 +0200 schrieb Alexandre Belloni:
Quoted 1 line
Hi,
[...]
Quoted 12 lines
Quoted 11 lines
So while we _might_ be able to get the PIO mode to work on the i.MX23 there is nothing in the doc stating that it's even meant to work. Even while PIO and DMA mode use the same internal mechanisms, there's still plenty of opportunities of fail in there. After all PIO mode relies on reading a debug register in the course of normal operation.
Only more extensive experimentation could show if we are in fact able to make it work, a first shot of using PIO mode on MX23 failed, so it might as well be that Juergens quick fix is correct and we have to disable PIO mode on MX23 altogether. That said please stop slapping the word "hack" over this patch until proven otherwise.
Quoted 5 lines
Still, Fabio and I have been experiencing the one second delay on i.mx28. From my experience, it appeared between 3.7 and 3.9, see:
http://www.spinics.net/lists/linux-i2c/msg11985.html. I still didn't take the time to investigate more yet but it confirms that PIO mode is working on i.mx28 though with an annoying delay.
Quoted 5 lines
Just for the record: I know that PIO mode is functional on i.MX28 because I have it working on one of our customer boards. I am rather much surprised it's still not working well for you, but haven't had a chance to investigate yet. The i.MX28 RM explicitly describes how to work with the PIO mode, so if it's not working this is bug somewhere.
I won't be surprised if there was more than one. The RM describes something , but has many flaws. I'm sure you noticed that yourself.
Quoted 4 lines
This thread/patch is about a whole different problem, namely PIO mode being non-functional on i.MX23. As the i.MX23 RM doesn't describe the PIO mode, there is no indication that it's even meant to work on this chip.
I'm not against the patch per-se, but can we not get the best of this and instead try some more to fix the PIO on MX23 so the MX23 users won't suffer performance degradation on already weak CPU?
Best regards,
Marek Vasut
PATCH 2/2 ARM: dts: imx28: Adjust i2c interrupt bindings
In reply to Lucas StachDear Lucas Stach,
Quoted 37 lines
Am Donnerstag, den 23.05.2013, 16:48 +0200 schrieb Marek Vasut:
Quoted 24 lines
Dear Juergen Beisert,
Quoted 13 lines
Hi Marek,
Quoted 10 lines · Marek Vasut
Quoted 10 lines
Quoted 1 line
i2c slowness is a different issue.
Same happens here for my i.M23 based platform. It seems the PIO mode does not work, or at least not like it works on a i.MX28. Each short transfer needs about one second (without an error message) but does not send anything on the I2C lines.
I need the following patches to make I2C master work within a 3.10-rc2 kernel:
Quoted 8 lines
Quoted 4 lines
Quoted 1 line
Subject: [PATCH] I2C/MXS: distinguish i.MX23 and i.MX28 based I2C
I'm all for it, but then ... won't it be better if you actually fixed the PIO and mixed-mode on MX23 instead of implementing such hack?
If the PIO mode or my patch is a hack depends on the point of view:
Lucas told me the PIO mode is *mentioned* but *not specified* in the i.MX23/i.MX28 datasheets.Quoted 8 lines
The PIO works the same way DMA does -- set up bits and then pump data into the DATA register.
Quoted 2 lines
So, the PIO mode seems to depend on some undocumented status bits in the i.MX28 I2C controller implementation.
How would DMA work then if it used undocumented registers ? It's in the documentation, just read it or ask FSL ;-)
While the PIO mode might use the same controller mechanisms as the DMA mode, PIO mode is _not_ a documented mode of operation for the i.MX23.
Quoted 8 lines
To quote the i.MX28 RM: "The I2C block on the i.MX28 supports a new PIO mode or soft-DMA mode.", which implies the PIO mode to be a new mode of operation not found on earlier i.MX SoCs.
The doc is slightly fuzzy here as the i.MX23 RM in contrary states:
"Short transmission (up to three bytes plus address) can be easily triggered using only PIO operations, i.e., no DMA setup required." But again it's not a documented mode of operation, i.MX23 doc only describes the DMA mode.
I'm CCing a Amaury Pouly so he can comment on this stuff. He did some hardware digging.
Quoted 5 lines
So while we _might_ be able to get the PIO mode to work on the i.MX23 there is nothing in the doc stating that it's even meant to work. Even while PIO and DMA mode use the same internal mechanisms, there's still plenty of opportunities of fail in there. After all PIO mode relies on reading a debug register in the course of normal operation.
What's the problem with that ? The debug register contains status bits.
Quoted 8 lines
Only more extensive experimentation could show if we are in fact able to make it work, a first shot of using PIO mode on MX23 failed, so it might as well be that Juergens quick fix is correct and we have to disable PIO mode on MX23 altogether. That said please stop slapping the word "hack" over this patch until proven otherwise.
Regards,
Lucas
Best regards,
Marek Vasut
Fabio Estevam
reviewerPATCH 2/2 ARM: dts: imx28: Adjust i2c interrupt bindings
In reply to Juergen BeisertHi Juergen,
Quoted 10 lines · Juergen Beisert
Hi Fabio,
Quoted 1 line
i2c slowness is a different issue.
Same happens here for my i.M23 based platform. It seems the PIO mode does not work, or at least not like it works on a i.MX28. Each short transfer needs about one second (without an error message) but does not send anything on the I2C lines.
I need the following patches to make I2C master work within a 3.10-rc2 kernel:
Thanks, I tried forcing dma mode and I got the following probe issue on a mx28evk running today's linux-next:
[ 6.403193] mxs-sgtl5000 sound.12: ASoC: CODEC (null) not registered [ 6.412175] mxs-sgtl5000 sound.12: snd_soc_register_card failed (-517) [ 6.425582] platform sound.12: Driver mxs-sgtl5000 requests probe deferral [ 6.434043] mmcblk0: p1 p2 p3 [ 6.439586] stmp3xxx-rtc 80056000.rtc: setting system clock to 2013-04-23 21:00:42 UTC (1366750842) [ 6.458141] mxs-sgtl5000 sound.12: ASoC: CODEC (null) not registered [ 6.464773] mxs-sgtl5000 sound.12: snd_soc_register_card failed (-517)
If I run linux-next with 2f769432d8554 reverted then I get a 23 seconds delay in the probe (1 second for each sgtl5000 register write).
Regards,
Fabio Estevam
PATCH 2/2 ARM: dts: imx28: Adjust i2c interrupt bindings
In reply to Juergen BeisertDear Juergen Beisert,
Quoted 9 lines
Hi Fabio,
Quoted 1 line
i2c slowness is a different issue.
Same happens here for my i.M23 based platform. It seems the PIO mode does not work, or at least not like it works on a i.MX28. Each short transfer needs about one second (without an error message) but does not send anything on the I2C lines.
Quoted 5 lines
I need the following patches to make I2C master work within a 3.10-rc2 kernel:
Subject: [PATCH] I2C/MXS: distinguish i.MX23 and i.MX28 based I2C controller Signed-off-by: Juergen Beisert <[email protected]>
Can you please properly submit this patch ?
Best regards,
Marek Vasut
PATCH 1/2 ARM: dts: imx: Remove custom dma-channel bindings
In reply to Fabio EstevamQuoted 4 lines · Fabio Estevam
Commit f30fb03 (ARM: dts: add generic DMA device tree binding for mxs-dma) added the generic DMA dt bindings, but missed to remove the old ones.
Remove the the old dma bindings.
I already had a patch [1] to remove those old DMA binding data (both channel and interrupt number) from client device nodes. And I'm waiting for mxs-saif/pcm to be converted to generic DMA binding helper to apply the patch. Mark has queued mxs-saif conversion for 3.11, and I have to wait for 3.11-rc1 to apply the dts cleanup.
Shawn
[1] http://thread.gmane.org/gmane.linux.alsa.devel/106027/focus=220815