Switching to 4.10

Switching to 4.10

[Joel Stanley]

Hello team,

I want to make the 4.10 change happen. We don't have the resource to
be maintaining two trees, and the 4.10 tree has a whole heap of
shinyness - faster network, faster flash access, improved I2C.

To this end I will stop merging patches in to 4.7. Please target the
4.10 tree from here on.

Use this thread to discuss concerns and coordinate the work that needs
to be done. Bits that I'm aware of

 - Eddie is updating userspace to the new FSI APIs
 - I need to merge the upstream FSI patchset
 - Brednan needs to post the lastest version of the i2c patchset which
I will merge
 - I need to merge benh's ftgmac100 improvements

Cheers,

Joel

Re: Switching to 4.10

[Brendan Higgins]

I just posted the latest version of the I2C driver this morning.

Cheers

Re: Switching to 4.10

[Benjamin Herrenschmidt]

On Mon, 2017-04-24 at 16:07 -0700, Brendan Higgins wrote:
> I just posted the latest version of the I2C driver this morning.

Thanks Brendan !

I'll try to review between flights :-)

Joel, can you and Andrew torture it on our systems (including UDC gunk
etc...) ?

Cheers,
Ben.

Re: Switching to 4.10

[Brendan Higgins]

BTW, I renamed the clock-frequency DT property to bus-frequency as
requested, so the *.dtsi will need to be updated.

Re: Switching to 4.10

[Xo Wang]

Hi Joel,

On Thu, Apr 20, 2017 at 7:30 PM, Joel Stanley <joel@jms.id.au> wrote:
> Hello team,
>
> I want to make the 4.10 change happen. We don't have the resource to
> be maintaining two trees, and the 4.10 tree has a whole heap of
> shinyness - faster network, faster flash access, improved I2C.
>
> To this end I will stop merging patches in to 4.7. Please target the
> 4.10 tree from here on.
>
> Use this thread to discuss concerns and coordinate the work that needs
> to be done. Bits that I'm aware of
>
>  - Eddie is updating userspace to the new FSI APIs
>  - I need to merge the upstream FSI patchset
>  - Brednan needs to post the lastest version of the i2c patchset which
> I will merge
>  - I need to merge benh's ftgmac100 improvements

We'll need to update the defconfig in the build system. Currently if I
compile the kernel and generate a new defconfig (make savedefconfig),
I see the following changes in the 4.10 generated defconfig:

@@ -73,7 +73,6 @@
 CONFIG_MTD_BLOCK=y
 CONFIG_MTD_PARTITIONED_MASTER=y
 CONFIG_MTD_SPI_NOR=y
-CONFIG_ASPEED_FLASH_SPI=y
 CONFIG_MTD_UBI=y
 CONFIG_MTD_UBI_FASTMAP=y
 CONFIG_MTD_UBI_BLOCK=y
@@ -116,7 +115,6 @@
 # CONFIG_SERIO is not set
 # CONFIG_VT is not set
 CONFIG_LEGACY_PTY_COUNT=4
-# CONFIG_DEVKMEM is not set
 CONFIG_SERIAL_8250=y
 # CONFIG_SERIAL_8250_DEPRECATED_OPTIONS is not set
 CONFIG_SERIAL_8250_CONSOLE=y
@@ -131,10 +129,8 @@
 CONFIG_I2C=y
 CONFIG_I2C_CHARDEV=y
 CONFIG_I2C_ASPEED=y
-CONFIG_SENSORS_MAX31785=y
 CONFIG_SENSORS_LM75=y
 CONFIG_SENSORS_NCT7904=y
-CONFIG_SENSORS_POWER8_OCC_I2C=y
 CONFIG_PMBUS=y
 CONFIG_SENSORS_ADM1275=y
 CONFIG_SENSORS_UCD9000=y
@@ -143,21 +139,14 @@
 # CONFIG_USB_SUPPORT is not set
 CONFIG_NEW_LEDS=y
 CONFIG_LEDS_CLASS=y
-CONFIG_LEDS_GPIO=y
 CONFIG_LEDS_TRIGGERS=y
 CONFIG_LEDS_TRIGGER_TIMER=y
 CONFIG_LEDS_TRIGGER_HEARTBEAT=y
-CONFIG_LEDS_TRIGGER_GPIO=y
 CONFIG_RTC_CLASS=y
 CONFIG_RTC_DRV_DS1307=y
 CONFIG_RTC_DRV_ASPEED=y
 CONFIG_MAILBOX=y
-CONFIG_ASPEED_MBOX=y
 # CONFIG_IOMMU_SUPPORT is not set
-CONFIG_FSI=y
-CONFIG_FSI_MASTER_GPIO=y
-CONFIG_FSI_SCOM=y
-CONFIG_FSI_I2C=y
 CONFIG_FIRMWARE_MEMMAP=y
 CONFIG_FANOTIFY=y
 CONFIG_AUTOFS4_FS=y

>
> Cheers,
>
> Joel

I know that CONFIG_ASPEED_FLASH_SPI is now CONFIG_SPI_ASPEED_SMC, but
I've yet to look at each of the other differences.

cheers
xo

Re: Switching to 4.10

[Xo Wang]

On Tue, Apr 25, 2017 at 4:57 PM, Xo Wang <xow@google.com> wrote:
> Hi Joel,
>
> On Thu, Apr 20, 2017 at 7:30 PM, Joel Stanley <joel@jms.id.au> wrote:
>> Hello team,
>>
>> I want to make the 4.10 change happen. We don't have the resource to
>> be maintaining two trees, and the 4.10 tree has a whole heap of
>> shinyness - faster network, faster flash access, improved I2C.
>>
>> To this end I will stop merging patches in to 4.7. Please target the
>> 4.10 tree from here on.
>>
>> Use this thread to discuss concerns and coordinate the work that needs
>> to be done. Bits that I'm aware of
>>
>>  - Eddie is updating userspace to the new FSI APIs
>>  - I need to merge the upstream FSI patchset
>>  - Brednan needs to post the lastest version of the i2c patchset which
>> I will merge
>>  - I need to merge benh's ftgmac100 improvements
>
> We'll need to update the defconfig in the build system. Currently if I
> compile the kernel and generate a new defconfig (make savedefconfig),
> I see the following changes in the 4.10 generated defconfig:
>
> @@ -73,7 +73,6 @@
>  CONFIG_MTD_BLOCK=y
>  CONFIG_MTD_PARTITIONED_MASTER=y
>  CONFIG_MTD_SPI_NOR=y
> -CONFIG_ASPEED_FLASH_SPI=y
>  CONFIG_MTD_UBI=y
>  CONFIG_MTD_UBI_FASTMAP=y
>  CONFIG_MTD_UBI_BLOCK=y
> @@ -116,7 +115,6 @@
>  # CONFIG_SERIO is not set
>  # CONFIG_VT is not set
>  CONFIG_LEGACY_PTY_COUNT=4
> -# CONFIG_DEVKMEM is not set
>  CONFIG_SERIAL_8250=y
>  # CONFIG_SERIAL_8250_DEPRECATED_OPTIONS is not set
>  CONFIG_SERIAL_8250_CONSOLE=y
> @@ -131,10 +129,8 @@
>  CONFIG_I2C=y
>  CONFIG_I2C_CHARDEV=y
>  CONFIG_I2C_ASPEED=y
> -CONFIG_SENSORS_MAX31785=y
>  CONFIG_SENSORS_LM75=y
>  CONFIG_SENSORS_NCT7904=y
> -CONFIG_SENSORS_POWER8_OCC_I2C=y
>  CONFIG_PMBUS=y
>  CONFIG_SENSORS_ADM1275=y
>  CONFIG_SENSORS_UCD9000=y
> @@ -143,21 +139,14 @@
>  # CONFIG_USB_SUPPORT is not set
>  CONFIG_NEW_LEDS=y
>  CONFIG_LEDS_CLASS=y
> -CONFIG_LEDS_GPIO=y
>  CONFIG_LEDS_TRIGGERS=y
>  CONFIG_LEDS_TRIGGER_TIMER=y
>  CONFIG_LEDS_TRIGGER_HEARTBEAT=y
> -CONFIG_LEDS_TRIGGER_GPIO=y
>  CONFIG_RTC_CLASS=y
>  CONFIG_RTC_DRV_DS1307=y
>  CONFIG_RTC_DRV_ASPEED=y
>  CONFIG_MAILBOX=y
> -CONFIG_ASPEED_MBOX=y
>  # CONFIG_IOMMU_SUPPORT is not set
> -CONFIG_FSI=y
> -CONFIG_FSI_MASTER_GPIO=y
> -CONFIG_FSI_SCOM=y
> -CONFIG_FSI_I2C=y
>  CONFIG_FIRMWARE_MEMMAP=y
>  CONFIG_FANOTIFY=y
>  CONFIG_AUTOFS4_FS=y
>
>>
>> Cheers,
>>
>> Joel
>
> I know that CONFIG_ASPEED_FLASH_SPI is now CONFIG_SPI_ASPEED_SMC, but
> I've yet to look at each of the other differences.
>
> cheers
> xo

OK, I guess it's not that much work to look at those configs. This
seems to work up to the limit of the missing FSI drivers as previously
mentioned:

--- a/meta-openbmc-bsp/meta-aspeed/meta-ast2500/recipes-kernel/linux/linux-obmc/defconfig
+++ b/meta-openbmc-bsp/meta-aspeed/meta-ast2500/recipes-kernel/linux/linux-obmc/defconfig
@@ -73,7 +73,7 @@ CONFIG_MTD_CMDLINE_PARTS=y
 CONFIG_MTD_BLOCK=y
 CONFIG_MTD_PARTITIONED_MASTER=y
 CONFIG_MTD_SPI_NOR=y
-CONFIG_ASPEED_FLASH_SPI=y
+CONFIG_SPI_ASPEED_SMC=y
 CONFIG_MTD_UBI=y
 CONFIG_MTD_UBI_FASTMAP=y
 CONFIG_MTD_UBI_BLOCK=y
@@ -116,7 +116,6 @@ CONFIG_INPUT_EVDEV=y
 # CONFIG_SERIO is not set
 # CONFIG_VT is not set
 CONFIG_LEGACY_PTY_COUNT=4
-# CONFIG_DEVKMEM is not set
 CONFIG_SERIAL_8250=y
 # CONFIG_SERIAL_8250_DEPRECATED_OPTIONS is not set
 CONFIG_SERIAL_8250_CONSOLE=y
@@ -131,10 +130,11 @@ CONFIG_ASPEED_BT_IPMI_BMC=y
 CONFIG_I2C=y
 CONFIG_I2C_CHARDEV=y
 CONFIG_I2C_ASPEED=y
-CONFIG_SENSORS_MAX31785=y
+CONFIG_GPIOLIB=y
+CONFIG_GPIO_SYSFS=y
+CONFIG_GPIO_ASPEED=y
 CONFIG_SENSORS_LM75=y
 CONFIG_SENSORS_NCT7904=y
-CONFIG_SENSORS_POWER8_OCC_I2C=y
 CONFIG_PMBUS=y
 CONFIG_SENSORS_ADM1275=y
 CONFIG_SENSORS_UCD9000=y
@@ -152,12 +152,8 @@ CONFIG_RTC_CLASS=y
 CONFIG_RTC_DRV_DS1307=y
 CONFIG_RTC_DRV_ASPEED=y
 CONFIG_MAILBOX=y
-CONFIG_ASPEED_MBOX=y
+CONFIG_ASPEED_LPC_MBOX=y
 # CONFIG_IOMMU_SUPPORT is not set
-CONFIG_FSI=y
-CONFIG_FSI_MASTER_GPIO=y
-CONFIG_FSI_SCOM=y
-CONFIG_FSI_I2C=y
 CONFIG_FIRMWARE_MEMMAP=y
 CONFIG_FANOTIFY=y
 CONFIG_AUTOFS4_FS=y

cheers
xo

Re: Switching to 4.10

[Rick Altherr]

[-- Attachment #1: Type: text/plain, Size: 4550 bytes --]

I saw Joel started https://gerrit.openbmc-project.xyz/#/c/2846/ which has a
defconfig for ast2400.  If you have one for ast2500, please upload a new
patch set to that change.

On Tue, Apr 25, 2017 at 5:30 PM, Xo Wang <xow@google.com> wrote:

> On Tue, Apr 25, 2017 at 4:57 PM, Xo Wang <xow@google.com> wrote:
> > Hi Joel,
> >
> > On Thu, Apr 20, 2017 at 7:30 PM, Joel Stanley <joel@jms.id.au> wrote:
> >> Hello team,
> >>
> >> I want to make the 4.10 change happen. We don't have the resource to
> >> be maintaining two trees, and the 4.10 tree has a whole heap of
> >> shinyness - faster network, faster flash access, improved I2C.
> >>
> >> To this end I will stop merging patches in to 4.7. Please target the
> >> 4.10 tree from here on.
> >>
> >> Use this thread to discuss concerns and coordinate the work that needs
> >> to be done. Bits that I'm aware of
> >>
> >>  - Eddie is updating userspace to the new FSI APIs
> >>  - I need to merge the upstream FSI patchset
> >>  - Brednan needs to post the lastest version of the i2c patchset which
> >> I will merge
> >>  - I need to merge benh's ftgmac100 improvements
> >
> > We'll need to update the defconfig in the build system. Currently if I
> > compile the kernel and generate a new defconfig (make savedefconfig),
> > I see the following changes in the 4.10 generated defconfig:
> >
> > @@ -73,7 +73,6 @@
> >  CONFIG_MTD_BLOCK=y
> >  CONFIG_MTD_PARTITIONED_MASTER=y
> >  CONFIG_MTD_SPI_NOR=y
> > -CONFIG_ASPEED_FLASH_SPI=y
> >  CONFIG_MTD_UBI=y
> >  CONFIG_MTD_UBI_FASTMAP=y
> >  CONFIG_MTD_UBI_BLOCK=y
> > @@ -116,7 +115,6 @@
> >  # CONFIG_SERIO is not set
> >  # CONFIG_VT is not set
> >  CONFIG_LEGACY_PTY_COUNT=4
> > -# CONFIG_DEVKMEM is not set
> >  CONFIG_SERIAL_8250=y
> >  # CONFIG_SERIAL_8250_DEPRECATED_OPTIONS is not set
> >  CONFIG_SERIAL_8250_CONSOLE=y
> > @@ -131,10 +129,8 @@
> >  CONFIG_I2C=y
> >  CONFIG_I2C_CHARDEV=y
> >  CONFIG_I2C_ASPEED=y
> > -CONFIG_SENSORS_MAX31785=y
> >  CONFIG_SENSORS_LM75=y
> >  CONFIG_SENSORS_NCT7904=y
> > -CONFIG_SENSORS_POWER8_OCC_I2C=y
> >  CONFIG_PMBUS=y
> >  CONFIG_SENSORS_ADM1275=y
> >  CONFIG_SENSORS_UCD9000=y
> > @@ -143,21 +139,14 @@
> >  # CONFIG_USB_SUPPORT is not set
> >  CONFIG_NEW_LEDS=y
> >  CONFIG_LEDS_CLASS=y
> > -CONFIG_LEDS_GPIO=y
> >  CONFIG_LEDS_TRIGGERS=y
> >  CONFIG_LEDS_TRIGGER_TIMER=y
> >  CONFIG_LEDS_TRIGGER_HEARTBEAT=y
> > -CONFIG_LEDS_TRIGGER_GPIO=y
> >  CONFIG_RTC_CLASS=y
> >  CONFIG_RTC_DRV_DS1307=y
> >  CONFIG_RTC_DRV_ASPEED=y
> >  CONFIG_MAILBOX=y
> > -CONFIG_ASPEED_MBOX=y
> >  # CONFIG_IOMMU_SUPPORT is not set
> > -CONFIG_FSI=y
> > -CONFIG_FSI_MASTER_GPIO=y
> > -CONFIG_FSI_SCOM=y
> > -CONFIG_FSI_I2C=y
> >  CONFIG_FIRMWARE_MEMMAP=y
> >  CONFIG_FANOTIFY=y
> >  CONFIG_AUTOFS4_FS=y
> >
> >>
> >> Cheers,
> >>
> >> Joel
> >
> > I know that CONFIG_ASPEED_FLASH_SPI is now CONFIG_SPI_ASPEED_SMC, but
> > I've yet to look at each of the other differences.
> >
> > cheers
> > xo
>
> OK, I guess it's not that much work to look at those configs. This
> seems to work up to the limit of the missing FSI drivers as previously
> mentioned:
>
> --- a/meta-openbmc-bsp/meta-aspeed/meta-ast2500/recipes-
> kernel/linux/linux-obmc/defconfig
> +++ b/meta-openbmc-bsp/meta-aspeed/meta-ast2500/recipes-
> kernel/linux/linux-obmc/defconfig
> @@ -73,7 +73,7 @@ CONFIG_MTD_CMDLINE_PARTS=y
>  CONFIG_MTD_BLOCK=y
>  CONFIG_MTD_PARTITIONED_MASTER=y
>  CONFIG_MTD_SPI_NOR=y
> -CONFIG_ASPEED_FLASH_SPI=y
> +CONFIG_SPI_ASPEED_SMC=y
>  CONFIG_MTD_UBI=y
>  CONFIG_MTD_UBI_FASTMAP=y
>  CONFIG_MTD_UBI_BLOCK=y
> @@ -116,7 +116,6 @@ CONFIG_INPUT_EVDEV=y
>  # CONFIG_SERIO is not set
>  # CONFIG_VT is not set
>  CONFIG_LEGACY_PTY_COUNT=4
> -# CONFIG_DEVKMEM is not set
>  CONFIG_SERIAL_8250=y
>  # CONFIG_SERIAL_8250_DEPRECATED_OPTIONS is not set
>  CONFIG_SERIAL_8250_CONSOLE=y
> @@ -131,10 +130,11 @@ CONFIG_ASPEED_BT_IPMI_BMC=y
>  CONFIG_I2C=y
>  CONFIG_I2C_CHARDEV=y
>  CONFIG_I2C_ASPEED=y
> -CONFIG_SENSORS_MAX31785=y
> +CONFIG_GPIOLIB=y
> +CONFIG_GPIO_SYSFS=y
> +CONFIG_GPIO_ASPEED=y
>  CONFIG_SENSORS_LM75=y
>  CONFIG_SENSORS_NCT7904=y
> -CONFIG_SENSORS_POWER8_OCC_I2C=y
>  CONFIG_PMBUS=y
>  CONFIG_SENSORS_ADM1275=y
>  CONFIG_SENSORS_UCD9000=y
> @@ -152,12 +152,8 @@ CONFIG_RTC_CLASS=y
>  CONFIG_RTC_DRV_DS1307=y
>  CONFIG_RTC_DRV_ASPEED=y
>  CONFIG_MAILBOX=y
> -CONFIG_ASPEED_MBOX=y
> +CONFIG_ASPEED_LPC_MBOX=y
>  # CONFIG_IOMMU_SUPPORT is not set
> -CONFIG_FSI=y
> -CONFIG_FSI_MASTER_GPIO=y
> -CONFIG_FSI_SCOM=y
> -CONFIG_FSI_I2C=y
>  CONFIG_FIRMWARE_MEMMAP=y
>  CONFIG_FANOTIFY=y
>  CONFIG_AUTOFS4_FS=y
>
> cheers
> xo
>

[-- Attachment #2: Type: text/html, Size: 6045 bytes --]

Re: Switching to 4.10

[Joel Stanley]

On Thu, Apr 27, 2017 at 3:00 AM, Rick Altherr <raltherr@google.com> wrote:
> I saw Joel started https://gerrit.openbmc-project.xyz/#/c/2846/ which has a
> defconfig for ast2400.  If you have one for ast2500, please upload a new
> patch set to that change.

I have defconfigs in the tree that are ready to go. I will upload them
to Gerrit when I get the chance.

I've pulled in:

 - Ben's ftgmac100 patches from the net-next tree
 - Brendan's i2c patches, v7
 - Andrew's GPIO and Pinconf patches from the gpio and pinconf trees
 - Chris Bostic's FSI series, v6

These are going through a test cycle on the ast2500evb and Plametto
tonight. Once I have an ack for the i2c device tree, I will add that
to the tree and push to github.

Eddie, can you let me know which version of your OCC hwmon patches you
want included. Same goes for the CFAM I2C patches. If you want to wait
for review then please ping your reviewers and we can see if we can
get the patches reviewed before we include them.

Brad, can you point me to the SBE FIFO work that you want included. If
you would prefer to wait for review that is okay.

Matt, I think do you want to merge the MAX31785 hwmon driver as-is for now?

Cheers,

JOel

>
> On Tue, Apr 25, 2017 at 5:30 PM, Xo Wang <xow@google.com> wrote:
>>
>> On Tue, Apr 25, 2017 at 4:57 PM, Xo Wang <xow@google.com> wrote:
>> > Hi Joel,
>> >
>> > On Thu, Apr 20, 2017 at 7:30 PM, Joel Stanley <joel@jms.id.au> wrote:
>> >> Hello team,
>> >>
>> >> I want to make the 4.10 change happen. We don't have the resource to
>> >> be maintaining two trees, and the 4.10 tree has a whole heap of
>> >> shinyness - faster network, faster flash access, improved I2C.
>> >>
>> >> To this end I will stop merging patches in to 4.7. Please target the
>> >> 4.10 tree from here on.
>> >>
>> >> Use this thread to discuss concerns and coordinate the work that needs
>> >> to be done. Bits that I'm aware of
>> >>
>> >>  - Eddie is updating userspace to the new FSI APIs
>> >>  - I need to merge the upstream FSI patchset
>> >>  - Brednan needs to post the lastest version of the i2c patchset which
>> >> I will merge
>> >>  - I need to merge benh's ftgmac100 improvements
>> >
>> > We'll need to update the defconfig in the build system. Currently if I
>> > compile the kernel and generate a new defconfig (make savedefconfig),
>> > I see the following changes in the 4.10 generated defconfig:
>> >
>> > @@ -73,7 +73,6 @@
>> >  CONFIG_MTD_BLOCK=y
>> >  CONFIG_MTD_PARTITIONED_MASTER=y
>> >  CONFIG_MTD_SPI_NOR=y
>> > -CONFIG_ASPEED_FLASH_SPI=y
>> >  CONFIG_MTD_UBI=y
>> >  CONFIG_MTD_UBI_FASTMAP=y
>> >  CONFIG_MTD_UBI_BLOCK=y
>> > @@ -116,7 +115,6 @@
>> >  # CONFIG_SERIO is not set
>> >  # CONFIG_VT is not set
>> >  CONFIG_LEGACY_PTY_COUNT=4
>> > -# CONFIG_DEVKMEM is not set
>> >  CONFIG_SERIAL_8250=y
>> >  # CONFIG_SERIAL_8250_DEPRECATED_OPTIONS is not set
>> >  CONFIG_SERIAL_8250_CONSOLE=y
>> > @@ -131,10 +129,8 @@
>> >  CONFIG_I2C=y
>> >  CONFIG_I2C_CHARDEV=y
>> >  CONFIG_I2C_ASPEED=y
>> > -CONFIG_SENSORS_MAX31785=y
>> >  CONFIG_SENSORS_LM75=y
>> >  CONFIG_SENSORS_NCT7904=y
>> > -CONFIG_SENSORS_POWER8_OCC_I2C=y
>> >  CONFIG_PMBUS=y
>> >  CONFIG_SENSORS_ADM1275=y
>> >  CONFIG_SENSORS_UCD9000=y
>> > @@ -143,21 +139,14 @@
>> >  # CONFIG_USB_SUPPORT is not set
>> >  CONFIG_NEW_LEDS=y
>> >  CONFIG_LEDS_CLASS=y
>> > -CONFIG_LEDS_GPIO=y
>> >  CONFIG_LEDS_TRIGGERS=y
>> >  CONFIG_LEDS_TRIGGER_TIMER=y
>> >  CONFIG_LEDS_TRIGGER_HEARTBEAT=y
>> > -CONFIG_LEDS_TRIGGER_GPIO=y
>> >  CONFIG_RTC_CLASS=y
>> >  CONFIG_RTC_DRV_DS1307=y
>> >  CONFIG_RTC_DRV_ASPEED=y
>> >  CONFIG_MAILBOX=y
>> > -CONFIG_ASPEED_MBOX=y
>> >  # CONFIG_IOMMU_SUPPORT is not set
>> > -CONFIG_FSI=y
>> > -CONFIG_FSI_MASTER_GPIO=y
>> > -CONFIG_FSI_SCOM=y
>> > -CONFIG_FSI_I2C=y
>> >  CONFIG_FIRMWARE_MEMMAP=y
>> >  CONFIG_FANOTIFY=y
>> >  CONFIG_AUTOFS4_FS=y
>> >
>> >>
>> >> Cheers,
>> >>
>> >> Joel
>> >
>> > I know that CONFIG_ASPEED_FLASH_SPI is now CONFIG_SPI_ASPEED_SMC, but
>> > I've yet to look at each of the other differences.
>> >
>> > cheers
>> > xo
>>
>> OK, I guess it's not that much work to look at those configs. This
>> seems to work up to the limit of the missing FSI drivers as previously
>> mentioned:
>>
>> ---
>> a/meta-openbmc-bsp/meta-aspeed/meta-ast2500/recipes-kernel/linux/linux-obmc/defconfig
>> +++
>> b/meta-openbmc-bsp/meta-aspeed/meta-ast2500/recipes-kernel/linux/linux-obmc/defconfig
>> @@ -73,7 +73,7 @@ CONFIG_MTD_CMDLINE_PARTS=y
>>  CONFIG_MTD_BLOCK=y
>>  CONFIG_MTD_PARTITIONED_MASTER=y
>>  CONFIG_MTD_SPI_NOR=y
>> -CONFIG_ASPEED_FLASH_SPI=y
>> +CONFIG_SPI_ASPEED_SMC=y
>>  CONFIG_MTD_UBI=y
>>  CONFIG_MTD_UBI_FASTMAP=y
>>  CONFIG_MTD_UBI_BLOCK=y
>> @@ -116,7 +116,6 @@ CONFIG_INPUT_EVDEV=y
>>  # CONFIG_SERIO is not set
>>  # CONFIG_VT is not set
>>  CONFIG_LEGACY_PTY_COUNT=4
>> -# CONFIG_DEVKMEM is not set
>>  CONFIG_SERIAL_8250=y
>>  # CONFIG_SERIAL_8250_DEPRECATED_OPTIONS is not set
>>  CONFIG_SERIAL_8250_CONSOLE=y
>> @@ -131,10 +130,11 @@ CONFIG_ASPEED_BT_IPMI_BMC=y
>>  CONFIG_I2C=y
>>  CONFIG_I2C_CHARDEV=y
>>  CONFIG_I2C_ASPEED=y
>> -CONFIG_SENSORS_MAX31785=y
>> +CONFIG_GPIOLIB=y
>> +CONFIG_GPIO_SYSFS=y
>> +CONFIG_GPIO_ASPEED=y
>>  CONFIG_SENSORS_LM75=y
>>  CONFIG_SENSORS_NCT7904=y
>> -CONFIG_SENSORS_POWER8_OCC_I2C=y
>>  CONFIG_PMBUS=y
>>  CONFIG_SENSORS_ADM1275=y
>>  CONFIG_SENSORS_UCD9000=y
>> @@ -152,12 +152,8 @@ CONFIG_RTC_CLASS=y
>>  CONFIG_RTC_DRV_DS1307=y
>>  CONFIG_RTC_DRV_ASPEED=y
>>  CONFIG_MAILBOX=y
>> -CONFIG_ASPEED_MBOX=y
>> +CONFIG_ASPEED_LPC_MBOX=y
>>  # CONFIG_IOMMU_SUPPORT is not set
>> -CONFIG_FSI=y
>> -CONFIG_FSI_MASTER_GPIO=y
>> -CONFIG_FSI_SCOM=y
>> -CONFIG_FSI_I2C=y
>>  CONFIG_FIRMWARE_MEMMAP=y
>>  CONFIG_FANOTIFY=y
>>  CONFIG_AUTOFS4_FS=y
>>
>> cheers
>> xo
>
>

Re: Switching to 4.10

[Joel Stanley]

On Thu, Apr 27, 2017 at 5:03 PM, Joel Stanley <joel@jms.id.au> wrote:
> On Thu, Apr 27, 2017 at 3:00 AM, Rick Altherr <raltherr@google.com> wrote:
>> I saw Joel started https://gerrit.openbmc-project.xyz/#/c/2846/ which has a
>> defconfig for ast2400.  If you have one for ast2500, please upload a new
>> patch set to that change.
>
> I have defconfigs in the tree that are ready to go. I will upload them
> to Gerrit when I get the chance.
>
> I've pulled in:
>
>  - Ben's ftgmac100 patches from the net-next tree
>  - Brendan's i2c patches, v7

The new i2c driver does not work with the model we have in Qemu. This
is a problem for testing in Qemu, but is not an issue on hardware:

 https://github.com/openbmc/qemu/issues/5

Cheers,

Joel

>  - Andrew's GPIO and Pinconf patches from the gpio and pinconf trees
>  - Chris Bostic's FSI series, v6
>
> These are going through a test cycle on the ast2500evb and Plametto
> tonight. Once I have an ack for the i2c device tree, I will add that
> to the tree and push to github.
>
> Eddie, can you let me know which version of your OCC hwmon patches you
> want included. Same goes for the CFAM I2C patches. If you want to wait
> for review then please ping your reviewers and we can see if we can
> get the patches reviewed before we include them.
>
> Brad, can you point me to the SBE FIFO work that you want included. If
> you would prefer to wait for review that is okay.
>
> Matt, I think do you want to merge the MAX31785 hwmon driver as-is for now?
>
> Cheers,
>
> JOel
>
>>
>> On Tue, Apr 25, 2017 at 5:30 PM, Xo Wang <xow@google.com> wrote:
>>>
>>> On Tue, Apr 25, 2017 at 4:57 PM, Xo Wang <xow@google.com> wrote:
>>> > Hi Joel,
>>> >
>>> > On Thu, Apr 20, 2017 at 7:30 PM, Joel Stanley <joel@jms.id.au> wrote:
>>> >> Hello team,
>>> >>
>>> >> I want to make the 4.10 change happen. We don't have the resource to
>>> >> be maintaining two trees, and the 4.10 tree has a whole heap of
>>> >> shinyness - faster network, faster flash access, improved I2C.
>>> >>
>>> >> To this end I will stop merging patches in to 4.7. Please target the
>>> >> 4.10 tree from here on.
>>> >>
>>> >> Use this thread to discuss concerns and coordinate the work that needs
>>> >> to be done. Bits that I'm aware of
>>> >>
>>> >>  - Eddie is updating userspace to the new FSI APIs
>>> >>  - I need to merge the upstream FSI patchset
>>> >>  - Brednan needs to post the lastest version of the i2c patchset which
>>> >> I will merge
>>> >>  - I need to merge benh's ftgmac100 improvements
>>> >
>>> > We'll need to update the defconfig in the build system. Currently if I
>>> > compile the kernel and generate a new defconfig (make savedefconfig),
>>> > I see the following changes in the 4.10 generated defconfig:
>>> >
>>> > @@ -73,7 +73,6 @@
>>> >  CONFIG_MTD_BLOCK=y
>>> >  CONFIG_MTD_PARTITIONED_MASTER=y
>>> >  CONFIG_MTD_SPI_NOR=y
>>> > -CONFIG_ASPEED_FLASH_SPI=y
>>> >  CONFIG_MTD_UBI=y
>>> >  CONFIG_MTD_UBI_FASTMAP=y
>>> >  CONFIG_MTD_UBI_BLOCK=y
>>> > @@ -116,7 +115,6 @@
>>> >  # CONFIG_SERIO is not set
>>> >  # CONFIG_VT is not set
>>> >  CONFIG_LEGACY_PTY_COUNT=4
>>> > -# CONFIG_DEVKMEM is not set
>>> >  CONFIG_SERIAL_8250=y
>>> >  # CONFIG_SERIAL_8250_DEPRECATED_OPTIONS is not set
>>> >  CONFIG_SERIAL_8250_CONSOLE=y
>>> > @@ -131,10 +129,8 @@
>>> >  CONFIG_I2C=y
>>> >  CONFIG_I2C_CHARDEV=y
>>> >  CONFIG_I2C_ASPEED=y
>>> > -CONFIG_SENSORS_MAX31785=y
>>> >  CONFIG_SENSORS_LM75=y
>>> >  CONFIG_SENSORS_NCT7904=y
>>> > -CONFIG_SENSORS_POWER8_OCC_I2C=y
>>> >  CONFIG_PMBUS=y
>>> >  CONFIG_SENSORS_ADM1275=y
>>> >  CONFIG_SENSORS_UCD9000=y
>>> > @@ -143,21 +139,14 @@
>>> >  # CONFIG_USB_SUPPORT is not set
>>> >  CONFIG_NEW_LEDS=y
>>> >  CONFIG_LEDS_CLASS=y
>>> > -CONFIG_LEDS_GPIO=y
>>> >  CONFIG_LEDS_TRIGGERS=y
>>> >  CONFIG_LEDS_TRIGGER_TIMER=y
>>> >  CONFIG_LEDS_TRIGGER_HEARTBEAT=y
>>> > -CONFIG_LEDS_TRIGGER_GPIO=y
>>> >  CONFIG_RTC_CLASS=y
>>> >  CONFIG_RTC_DRV_DS1307=y
>>> >  CONFIG_RTC_DRV_ASPEED=y
>>> >  CONFIG_MAILBOX=y
>>> > -CONFIG_ASPEED_MBOX=y
>>> >  # CONFIG_IOMMU_SUPPORT is not set
>>> > -CONFIG_FSI=y
>>> > -CONFIG_FSI_MASTER_GPIO=y
>>> > -CONFIG_FSI_SCOM=y
>>> > -CONFIG_FSI_I2C=y
>>> >  CONFIG_FIRMWARE_MEMMAP=y
>>> >  CONFIG_FANOTIFY=y
>>> >  CONFIG_AUTOFS4_FS=y
>>> >
>>> >>
>>> >> Cheers,
>>> >>
>>> >> Joel
>>> >
>>> > I know that CONFIG_ASPEED_FLASH_SPI is now CONFIG_SPI_ASPEED_SMC, but
>>> > I've yet to look at each of the other differences.
>>> >
>>> > cheers
>>> > xo
>>>
>>> OK, I guess it's not that much work to look at those configs. This
>>> seems to work up to the limit of the missing FSI drivers as previously
>>> mentioned:
>>>
>>> ---
>>> a/meta-openbmc-bsp/meta-aspeed/meta-ast2500/recipes-kernel/linux/linux-obmc/defconfig
>>> +++
>>> b/meta-openbmc-bsp/meta-aspeed/meta-ast2500/recipes-kernel/linux/linux-obmc/defconfig
>>> @@ -73,7 +73,7 @@ CONFIG_MTD_CMDLINE_PARTS=y
>>>  CONFIG_MTD_BLOCK=y
>>>  CONFIG_MTD_PARTITIONED_MASTER=y
>>>  CONFIG_MTD_SPI_NOR=y
>>> -CONFIG_ASPEED_FLASH_SPI=y
>>> +CONFIG_SPI_ASPEED_SMC=y
>>>  CONFIG_MTD_UBI=y
>>>  CONFIG_MTD_UBI_FASTMAP=y
>>>  CONFIG_MTD_UBI_BLOCK=y
>>> @@ -116,7 +116,6 @@ CONFIG_INPUT_EVDEV=y
>>>  # CONFIG_SERIO is not set
>>>  # CONFIG_VT is not set
>>>  CONFIG_LEGACY_PTY_COUNT=4
>>> -# CONFIG_DEVKMEM is not set
>>>  CONFIG_SERIAL_8250=y
>>>  # CONFIG_SERIAL_8250_DEPRECATED_OPTIONS is not set
>>>  CONFIG_SERIAL_8250_CONSOLE=y
>>> @@ -131,10 +130,11 @@ CONFIG_ASPEED_BT_IPMI_BMC=y
>>>  CONFIG_I2C=y
>>>  CONFIG_I2C_CHARDEV=y
>>>  CONFIG_I2C_ASPEED=y
>>> -CONFIG_SENSORS_MAX31785=y
>>> +CONFIG_GPIOLIB=y
>>> +CONFIG_GPIO_SYSFS=y
>>> +CONFIG_GPIO_ASPEED=y
>>>  CONFIG_SENSORS_LM75=y
>>>  CONFIG_SENSORS_NCT7904=y
>>> -CONFIG_SENSORS_POWER8_OCC_I2C=y
>>>  CONFIG_PMBUS=y
>>>  CONFIG_SENSORS_ADM1275=y
>>>  CONFIG_SENSORS_UCD9000=y
>>> @@ -152,12 +152,8 @@ CONFIG_RTC_CLASS=y
>>>  CONFIG_RTC_DRV_DS1307=y
>>>  CONFIG_RTC_DRV_ASPEED=y
>>>  CONFIG_MAILBOX=y
>>> -CONFIG_ASPEED_MBOX=y
>>> +CONFIG_ASPEED_LPC_MBOX=y
>>>  # CONFIG_IOMMU_SUPPORT is not set
>>> -CONFIG_FSI=y
>>> -CONFIG_FSI_MASTER_GPIO=y
>>> -CONFIG_FSI_SCOM=y
>>> -CONFIG_FSI_I2C=y
>>>  CONFIG_FIRMWARE_MEMMAP=y
>>>  CONFIG_FANOTIFY=y
>>>  CONFIG_AUTOFS4_FS=y
>>>
>>> cheers
>>> xo
>>
>>

Re: Switching to 4.10

[Matthew Barth]

On 04/27/17 2:33 AM, Joel Stanley wrote:
> On Thu, Apr 27, 2017 at 3:00 AM, Rick Altherr <raltherr@google.com> wrote:
>> I saw Joel started https://gerrit.openbmc-project.xyz/#/c/2846/ which has a
>> defconfig for ast2400.  If you have one for ast2500, please upload a new
>> patch set to that change.
> I have defconfigs in the tree that are ready to go. I will upload them
> to Gerrit when I get the chance.
>
> I've pulled in:
>
>   - Ben's ftgmac100 patches from the net-next tree
>   - Brendan's i2c patches, v7
>   - Andrew's GPIO and Pinconf patches from the gpio and pinconf trees
>   - Chris Bostic's FSI series, v6
>
> These are going through a test cycle on the ast2500evb and Plametto
> tonight. Once I have an ack for the i2c device tree, I will add that
> to the tree and push to github.
>
> Eddie, can you let me know which version of your OCC hwmon patches you
> want included. Same goes for the CFAM I2C patches. If you want to wait
> for review then please ping your reviewers and we can see if we can
> get the patches reviewed before we include them.
>
> Brad, can you point me to the SBE FIFO work that you want included. If
> you would prefer to wait for review that is okay.
>
> Matt, I think do you want to merge the MAX31785 hwmon driver as-is for now?
I thought we already have a MAX31785 driver merged, so what is 
'as-is'...does that include the update supporting reading the tach 
feedback from both rotors as described in 
issue#652(https://github.com/openbmc/openbmc/issues/652)? The commit you 
referenced in that issue does not have the dual tach feedback support yet.

At this point in time, it's important we get the MAX31785 driver updated 
with this support along with the PCA9552 driver presence & led 
verification of the fan enclosures.

Thanks,

Matt
>
> Cheers,
>
> JOel
>
>> On Tue, Apr 25, 2017 at 5:30 PM, Xo Wang <xow@google.com> wrote:
>>> On Tue, Apr 25, 2017 at 4:57 PM, Xo Wang <xow@google.com> wrote:
>>>> Hi Joel,
>>>>
>>>> On Thu, Apr 20, 2017 at 7:30 PM, Joel Stanley <joel@jms.id.au> wrote:
>>>>> Hello team,
>>>>>
>>>>> I want to make the 4.10 change happen. We don't have the resource to
>>>>> be maintaining two trees, and the 4.10 tree has a whole heap of
>>>>> shinyness - faster network, faster flash access, improved I2C.
>>>>>
>>>>> To this end I will stop merging patches in to 4.7. Please target the
>>>>> 4.10 tree from here on.
>>>>>
>>>>> Use this thread to discuss concerns and coordinate the work that needs
>>>>> to be done. Bits that I'm aware of
>>>>>
>>>>>   - Eddie is updating userspace to the new FSI APIs
>>>>>   - I need to merge the upstream FSI patchset
>>>>>   - Brednan needs to post the lastest version of the i2c patchset which
>>>>> I will merge
>>>>>   - I need to merge benh's ftgmac100 improvements
>>>> We'll need to update the defconfig in the build system. Currently if I
>>>> compile the kernel and generate a new defconfig (make savedefconfig),
>>>> I see the following changes in the 4.10 generated defconfig:
>>>>
>>>> @@ -73,7 +73,6 @@
>>>>   CONFIG_MTD_BLOCK=y
>>>>   CONFIG_MTD_PARTITIONED_MASTER=y
>>>>   CONFIG_MTD_SPI_NOR=y
>>>> -CONFIG_ASPEED_FLASH_SPI=y
>>>>   CONFIG_MTD_UBI=y
>>>>   CONFIG_MTD_UBI_FASTMAP=y
>>>>   CONFIG_MTD_UBI_BLOCK=y
>>>> @@ -116,7 +115,6 @@
>>>>   # CONFIG_SERIO is not set
>>>>   # CONFIG_VT is not set
>>>>   CONFIG_LEGACY_PTY_COUNT=4
>>>> -# CONFIG_DEVKMEM is not set
>>>>   CONFIG_SERIAL_8250=y
>>>>   # CONFIG_SERIAL_8250_DEPRECATED_OPTIONS is not set
>>>>   CONFIG_SERIAL_8250_CONSOLE=y
>>>> @@ -131,10 +129,8 @@
>>>>   CONFIG_I2C=y
>>>>   CONFIG_I2C_CHARDEV=y
>>>>   CONFIG_I2C_ASPEED=y
>>>> -CONFIG_SENSORS_MAX31785=y
>>>>   CONFIG_SENSORS_LM75=y
>>>>   CONFIG_SENSORS_NCT7904=y
>>>> -CONFIG_SENSORS_POWER8_OCC_I2C=y
>>>>   CONFIG_PMBUS=y
>>>>   CONFIG_SENSORS_ADM1275=y
>>>>   CONFIG_SENSORS_UCD9000=y
>>>> @@ -143,21 +139,14 @@
>>>>   # CONFIG_USB_SUPPORT is not set
>>>>   CONFIG_NEW_LEDS=y
>>>>   CONFIG_LEDS_CLASS=y
>>>> -CONFIG_LEDS_GPIO=y
>>>>   CONFIG_LEDS_TRIGGERS=y
>>>>   CONFIG_LEDS_TRIGGER_TIMER=y
>>>>   CONFIG_LEDS_TRIGGER_HEARTBEAT=y
>>>> -CONFIG_LEDS_TRIGGER_GPIO=y
>>>>   CONFIG_RTC_CLASS=y
>>>>   CONFIG_RTC_DRV_DS1307=y
>>>>   CONFIG_RTC_DRV_ASPEED=y
>>>>   CONFIG_MAILBOX=y
>>>> -CONFIG_ASPEED_MBOX=y
>>>>   # CONFIG_IOMMU_SUPPORT is not set
>>>> -CONFIG_FSI=y
>>>> -CONFIG_FSI_MASTER_GPIO=y
>>>> -CONFIG_FSI_SCOM=y
>>>> -CONFIG_FSI_I2C=y
>>>>   CONFIG_FIRMWARE_MEMMAP=y
>>>>   CONFIG_FANOTIFY=y
>>>>   CONFIG_AUTOFS4_FS=y
>>>>
>>>>> Cheers,
>>>>>
>>>>> Joel
>>>> I know that CONFIG_ASPEED_FLASH_SPI is now CONFIG_SPI_ASPEED_SMC, but
>>>> I've yet to look at each of the other differences.
>>>>
>>>> cheers
>>>> xo
>>> OK, I guess it's not that much work to look at those configs. This
>>> seems to work up to the limit of the missing FSI drivers as previously
>>> mentioned:
>>>
>>> ---
>>> a/meta-openbmc-bsp/meta-aspeed/meta-ast2500/recipes-kernel/linux/linux-obmc/defconfig
>>> +++
>>> b/meta-openbmc-bsp/meta-aspeed/meta-ast2500/recipes-kernel/linux/linux-obmc/defconfig
>>> @@ -73,7 +73,7 @@ CONFIG_MTD_CMDLINE_PARTS=y
>>>   CONFIG_MTD_BLOCK=y
>>>   CONFIG_MTD_PARTITIONED_MASTER=y
>>>   CONFIG_MTD_SPI_NOR=y
>>> -CONFIG_ASPEED_FLASH_SPI=y
>>> +CONFIG_SPI_ASPEED_SMC=y
>>>   CONFIG_MTD_UBI=y
>>>   CONFIG_MTD_UBI_FASTMAP=y
>>>   CONFIG_MTD_UBI_BLOCK=y
>>> @@ -116,7 +116,6 @@ CONFIG_INPUT_EVDEV=y
>>>   # CONFIG_SERIO is not set
>>>   # CONFIG_VT is not set
>>>   CONFIG_LEGACY_PTY_COUNT=4
>>> -# CONFIG_DEVKMEM is not set
>>>   CONFIG_SERIAL_8250=y
>>>   # CONFIG_SERIAL_8250_DEPRECATED_OPTIONS is not set
>>>   CONFIG_SERIAL_8250_CONSOLE=y
>>> @@ -131,10 +130,11 @@ CONFIG_ASPEED_BT_IPMI_BMC=y
>>>   CONFIG_I2C=y
>>>   CONFIG_I2C_CHARDEV=y
>>>   CONFIG_I2C_ASPEED=y
>>> -CONFIG_SENSORS_MAX31785=y
>>> +CONFIG_GPIOLIB=y
>>> +CONFIG_GPIO_SYSFS=y
>>> +CONFIG_GPIO_ASPEED=y
>>>   CONFIG_SENSORS_LM75=y
>>>   CONFIG_SENSORS_NCT7904=y
>>> -CONFIG_SENSORS_POWER8_OCC_I2C=y
>>>   CONFIG_PMBUS=y
>>>   CONFIG_SENSORS_ADM1275=y
>>>   CONFIG_SENSORS_UCD9000=y
>>> @@ -152,12 +152,8 @@ CONFIG_RTC_CLASS=y
>>>   CONFIG_RTC_DRV_DS1307=y
>>>   CONFIG_RTC_DRV_ASPEED=y
>>>   CONFIG_MAILBOX=y
>>> -CONFIG_ASPEED_MBOX=y
>>> +CONFIG_ASPEED_LPC_MBOX=y
>>>   # CONFIG_IOMMU_SUPPORT is not set
>>> -CONFIG_FSI=y
>>> -CONFIG_FSI_MASTER_GPIO=y
>>> -CONFIG_FSI_SCOM=y
>>> -CONFIG_FSI_I2C=y
>>>   CONFIG_FIRMWARE_MEMMAP=y
>>>   CONFIG_FANOTIFY=y
>>>   CONFIG_AUTOFS4_FS=y
>>>
>>> cheers
>>> xo
>>

Re: Switching to 4.10

[Rick Altherr]

[-- Attachment #1: Type: text/plain, Size: 6334 bytes --]

On Thu, Apr 27, 2017 at 12:33 AM, Joel Stanley <joel@jms.id.au> wrote:

> On Thu, Apr 27, 2017 at 3:00 AM, Rick Altherr <raltherr@google.com> wrote:
> > I saw Joel started https://gerrit.openbmc-project.xyz/#/c/2846/ which
> has a
> > defconfig for ast2400.  If you have one for ast2500, please upload a new
> > patch set to that change.
>
> I have defconfigs in the tree that are ready to go. I will upload them
> to Gerrit when I get the chance.
>
>
I've been holding off on testing some RNG changes because it depends on
4.10.  Can you please upload these defconfigs so I can build a full 4.10
image more easily?


> I've pulled in:
>
>  - Ben's ftgmac100 patches from the net-next tree
>  - Brendan's i2c patches, v7
>  - Andrew's GPIO and Pinconf patches from the gpio and pinconf trees
>  - Chris Bostic's FSI series, v6
>
> These are going through a test cycle on the ast2500evb and Plametto
> tonight. Once I have an ack for the i2c device tree, I will add that
> to the tree and push to github.
>
> Eddie, can you let me know which version of your OCC hwmon patches you
> want included. Same goes for the CFAM I2C patches. If you want to wait
> for review then please ping your reviewers and we can see if we can
> get the patches reviewed before we include them.
>
> Brad, can you point me to the SBE FIFO work that you want included. If
> you would prefer to wait for review that is okay.
>
> Matt, I think do you want to merge the MAX31785 hwmon driver as-is for now?
>
> Cheers,
>
> JOel
>
> >
> > On Tue, Apr 25, 2017 at 5:30 PM, Xo Wang <xow@google.com> wrote:
> >>
> >> On Tue, Apr 25, 2017 at 4:57 PM, Xo Wang <xow@google.com> wrote:
> >> > Hi Joel,
> >> >
> >> > On Thu, Apr 20, 2017 at 7:30 PM, Joel Stanley <joel@jms.id.au> wrote:
> >> >> Hello team,
> >> >>
> >> >> I want to make the 4.10 change happen. We don't have the resource to
> >> >> be maintaining two trees, and the 4.10 tree has a whole heap of
> >> >> shinyness - faster network, faster flash access, improved I2C.
> >> >>
> >> >> To this end I will stop merging patches in to 4.7. Please target the
> >> >> 4.10 tree from here on.
> >> >>
> >> >> Use this thread to discuss concerns and coordinate the work that
> needs
> >> >> to be done. Bits that I'm aware of
> >> >>
> >> >>  - Eddie is updating userspace to the new FSI APIs
> >> >>  - I need to merge the upstream FSI patchset
> >> >>  - Brednan needs to post the lastest version of the i2c patchset
> which
> >> >> I will merge
> >> >>  - I need to merge benh's ftgmac100 improvements
> >> >
> >> > We'll need to update the defconfig in the build system. Currently if I
> >> > compile the kernel and generate a new defconfig (make savedefconfig),
> >> > I see the following changes in the 4.10 generated defconfig:
> >> >
> >> > @@ -73,7 +73,6 @@
> >> >  CONFIG_MTD_BLOCK=y
> >> >  CONFIG_MTD_PARTITIONED_MASTER=y
> >> >  CONFIG_MTD_SPI_NOR=y
> >> > -CONFIG_ASPEED_FLASH_SPI=y
> >> >  CONFIG_MTD_UBI=y
> >> >  CONFIG_MTD_UBI_FASTMAP=y
> >> >  CONFIG_MTD_UBI_BLOCK=y
> >> > @@ -116,7 +115,6 @@
> >> >  # CONFIG_SERIO is not set
> >> >  # CONFIG_VT is not set
> >> >  CONFIG_LEGACY_PTY_COUNT=4
> >> > -# CONFIG_DEVKMEM is not set
> >> >  CONFIG_SERIAL_8250=y
> >> >  # CONFIG_SERIAL_8250_DEPRECATED_OPTIONS is not set
> >> >  CONFIG_SERIAL_8250_CONSOLE=y
> >> > @@ -131,10 +129,8 @@
> >> >  CONFIG_I2C=y
> >> >  CONFIG_I2C_CHARDEV=y
> >> >  CONFIG_I2C_ASPEED=y
> >> > -CONFIG_SENSORS_MAX31785=y
> >> >  CONFIG_SENSORS_LM75=y
> >> >  CONFIG_SENSORS_NCT7904=y
> >> > -CONFIG_SENSORS_POWER8_OCC_I2C=y
> >> >  CONFIG_PMBUS=y
> >> >  CONFIG_SENSORS_ADM1275=y
> >> >  CONFIG_SENSORS_UCD9000=y
> >> > @@ -143,21 +139,14 @@
> >> >  # CONFIG_USB_SUPPORT is not set
> >> >  CONFIG_NEW_LEDS=y
> >> >  CONFIG_LEDS_CLASS=y
> >> > -CONFIG_LEDS_GPIO=y
> >> >  CONFIG_LEDS_TRIGGERS=y
> >> >  CONFIG_LEDS_TRIGGER_TIMER=y
> >> >  CONFIG_LEDS_TRIGGER_HEARTBEAT=y
> >> > -CONFIG_LEDS_TRIGGER_GPIO=y
> >> >  CONFIG_RTC_CLASS=y
> >> >  CONFIG_RTC_DRV_DS1307=y
> >> >  CONFIG_RTC_DRV_ASPEED=y
> >> >  CONFIG_MAILBOX=y
> >> > -CONFIG_ASPEED_MBOX=y
> >> >  # CONFIG_IOMMU_SUPPORT is not set
> >> > -CONFIG_FSI=y
> >> > -CONFIG_FSI_MASTER_GPIO=y
> >> > -CONFIG_FSI_SCOM=y
> >> > -CONFIG_FSI_I2C=y
> >> >  CONFIG_FIRMWARE_MEMMAP=y
> >> >  CONFIG_FANOTIFY=y
> >> >  CONFIG_AUTOFS4_FS=y
> >> >
> >> >>
> >> >> Cheers,
> >> >>
> >> >> Joel
> >> >
> >> > I know that CONFIG_ASPEED_FLASH_SPI is now CONFIG_SPI_ASPEED_SMC, but
> >> > I've yet to look at each of the other differences.
> >> >
> >> > cheers
> >> > xo
> >>
> >> OK, I guess it's not that much work to look at those configs. This
> >> seems to work up to the limit of the missing FSI drivers as previously
> >> mentioned:
> >>
> >> ---
> >> a/meta-openbmc-bsp/meta-aspeed/meta-ast2500/recipes-
> kernel/linux/linux-obmc/defconfig
> >> +++
> >> b/meta-openbmc-bsp/meta-aspeed/meta-ast2500/recipes-
> kernel/linux/linux-obmc/defconfig
> >> @@ -73,7 +73,7 @@ CONFIG_MTD_CMDLINE_PARTS=y
> >>  CONFIG_MTD_BLOCK=y
> >>  CONFIG_MTD_PARTITIONED_MASTER=y
> >>  CONFIG_MTD_SPI_NOR=y
> >> -CONFIG_ASPEED_FLASH_SPI=y
> >> +CONFIG_SPI_ASPEED_SMC=y
> >>  CONFIG_MTD_UBI=y
> >>  CONFIG_MTD_UBI_FASTMAP=y
> >>  CONFIG_MTD_UBI_BLOCK=y
> >> @@ -116,7 +116,6 @@ CONFIG_INPUT_EVDEV=y
> >>  # CONFIG_SERIO is not set
> >>  # CONFIG_VT is not set
> >>  CONFIG_LEGACY_PTY_COUNT=4
> >> -# CONFIG_DEVKMEM is not set
> >>  CONFIG_SERIAL_8250=y
> >>  # CONFIG_SERIAL_8250_DEPRECATED_OPTIONS is not set
> >>  CONFIG_SERIAL_8250_CONSOLE=y
> >> @@ -131,10 +130,11 @@ CONFIG_ASPEED_BT_IPMI_BMC=y
> >>  CONFIG_I2C=y
> >>  CONFIG_I2C_CHARDEV=y
> >>  CONFIG_I2C_ASPEED=y
> >> -CONFIG_SENSORS_MAX31785=y
> >> +CONFIG_GPIOLIB=y
> >> +CONFIG_GPIO_SYSFS=y
> >> +CONFIG_GPIO_ASPEED=y
> >>  CONFIG_SENSORS_LM75=y
> >>  CONFIG_SENSORS_NCT7904=y
> >> -CONFIG_SENSORS_POWER8_OCC_I2C=y
> >>  CONFIG_PMBUS=y
> >>  CONFIG_SENSORS_ADM1275=y
> >>  CONFIG_SENSORS_UCD9000=y
> >> @@ -152,12 +152,8 @@ CONFIG_RTC_CLASS=y
> >>  CONFIG_RTC_DRV_DS1307=y
> >>  CONFIG_RTC_DRV_ASPEED=y
> >>  CONFIG_MAILBOX=y
> >> -CONFIG_ASPEED_MBOX=y
> >> +CONFIG_ASPEED_LPC_MBOX=y
> >>  # CONFIG_IOMMU_SUPPORT is not set
> >> -CONFIG_FSI=y
> >> -CONFIG_FSI_MASTER_GPIO=y
> >> -CONFIG_FSI_SCOM=y
> >> -CONFIG_FSI_I2C=y
> >>  CONFIG_FIRMWARE_MEMMAP=y
> >>  CONFIG_FANOTIFY=y
> >>  CONFIG_AUTOFS4_FS=y
> >>
> >> cheers
> >> xo
> >
> >
>

[-- Attachment #2: Type: text/html, Size: 9044 bytes --]

Re: Switching to 4.10

[Edward James]

[-- Attachment #1: Type: text/html, Size: 9642 bytes --]

Re: Switching to 4.10

[Joel Stanley]

On Fri, Apr 28, 2017 at 1:39 AM, Matthew Barth
<msbarth@linux.vnet.ibm.com> wrote:
>
>
> On 04/27/17 2:33 AM, Joel Stanley wrote:
>>
>> On Thu, Apr 27, 2017 at 3:00 AM, Rick Altherr <raltherr@google.com> wrote:
>>>
>>> I saw Joel started https://gerrit.openbmc-project.xyz/#/c/2846/ which has
>>> a
>>> defconfig for ast2400.  If you have one for ast2500, please upload a new
>>> patch set to that change.
>>
>> I have defconfigs in the tree that are ready to go. I will upload them
>> to Gerrit when I get the chance.
>>
>> I've pulled in:
>>
>>   - Ben's ftgmac100 patches from the net-next tree
>>   - Brendan's i2c patches, v7
>>   - Andrew's GPIO and Pinconf patches from the gpio and pinconf trees
>>   - Chris Bostic's FSI series, v6
>>
>> These are going through a test cycle on the ast2500evb and Plametto
>> tonight. Once I have an ack for the i2c device tree, I will add that
>> to the tree and push to github.
>>
>> Eddie, can you let me know which version of your OCC hwmon patches you
>> want included. Same goes for the CFAM I2C patches. If you want to wait
>> for review then please ping your reviewers and we can see if we can
>> get the patches reviewed before we include them.
>>
>> Brad, can you point me to the SBE FIFO work that you want included. If
>> you would prefer to wait for review that is okay.
>>
>> Matt, I think do you want to merge the MAX31785 hwmon driver as-is for
>> now?
>
> I thought we already have a MAX31785 driver merged, so what is
> 'as-is'...does that include the update supporting reading the tach feedback
> from both rotors as described in
> issue#652(https://github.com/openbmc/openbmc/issues/652)? The commit you
> referenced in that issue does not have the dual tach feedback support yet.

No, that driver is not upstream and therefore it is not in the tree.
It needs to be included in the OpenBMC tree as part of this work.

Thanks for clarifying the issue about dual tach feedback.

Is it useful for you to have the driver merged as-is, or does it need
to have that feature enabled before you can do any testing?

Cheers,

Joel

>
> At this point in time, it's important we get the MAX31785 driver updated
> with this support along with the PCA9552 driver presence & led verification
> of the fan enclosures.
>

Re: Switching to 4.10

[Joel Stanley]

On Fri, Apr 28, 2017 at 7:24 AM, Edward James <eajames@us.ibm.com> wrote:
> Hey Joel,
>
> I don't think we can pull in the occ-hwmon driver yet, because there will
> some moderate restructuring due to the way the SBEFIFO and OCC driver
> interact for P9, which I didn't anticipate in how I wrote the P8 side of
> things. Haven't gotten to that yet.

Thanks for the update. Keep me in the loop.

> For the I2C driver, I have the new kernel I2C FSI client driver on the obmc
> mailing list, but need some feedback on the dt bindings. The old FSP-based
> I2C driver can be put in, but needs some work for the endianness changes,
> which I'm hoping to get done tomorrow/monday.

I would like to avoid pulling in the FSP i2c driver. Can you chat with
Jk and Brad to get the bindings sorted, and we can use that one
instead?

Cheers,

Joel

>
> Thanks,
> Eddie
>
>
> ----- Original message -----
> From: Joel Stanley <joel@jms.id.au>
> Sent by: joel.stan@gmail.com
> To: Rick Altherr <raltherr@google.com>, Brad Bishop
> <bradleyb@fuzziesquirrel.com>, msbarth <msbarth@linux.vnet.ibm.com>, Edward
> James <eajames@us.ibm.com>
> Cc: Xo Wang <xow@google.com>, OpenBMC Maillist <openbmc@lists.ozlabs.org>,
> Brendan Higgins <brendanhiggins@google.com>
> Subject: Re: Switching to 4.10
> Date: Thu, Apr 27, 2017 2:33 AM
>
> On Thu, Apr 27, 2017 at 3:00 AM, Rick Altherr <raltherr@google.com> wrote:
>> I saw Joel started https://gerrit.openbmc-project.xyz/#/c/2846/ which has
>> a
>> defconfig for ast2400.  If you have one for ast2500, please upload a new
>> patch set to that change.
>
> I have defconfigs in the tree that are ready to go. I will upload them
> to Gerrit when I get the chance.
>
> I've pulled in:
>
>  - Ben's ftgmac100 patches from the net-next tree
>  - Brendan's i2c patches, v7
>  - Andrew's GPIO and Pinconf patches from the gpio and pinconf trees
>  - Chris Bostic's FSI series, v6
>
> These are going through a test cycle on the ast2500evb and Plametto
> tonight. Once I have an ack for the i2c device tree, I will add that
> to the tree and push to github.
>
> Eddie, can you let me know which version of your OCC hwmon patches you
> want included. Same goes for the CFAM I2C patches. If you want to wait
> for review then please ping your reviewers and we can see if we can
> get the patches reviewed before we include them.
>
> Brad, can you point me to the SBE FIFO work that you want included. If
> you would prefer to wait for review that is okay.
>
> Matt, I think do you want to merge the MAX31785 hwmon driver as-is for now?
>
> Cheers,
>
> JOel
>
>>
>> On Tue, Apr 25, 2017 at 5:30 PM, Xo Wang <xow@google.com> wrote:
>>>
>>> On Tue, Apr 25, 2017 at 4:57 PM, Xo Wang <xow@google.com> wrote:
>>> > Hi Joel,
>>> >
>>> > On Thu, Apr 20, 2017 at 7:30 PM, Joel Stanley <joel@jms.id.au> wrote:
>>> >> Hello team,
>>> >>
>>> >> I want to make the 4.10 change happen. We don't have the resource to
>>> >> be maintaining two trees, and the 4.10 tree has a whole heap of
>>> >> shinyness - faster network, faster flash access, improved I2C.
>>> >>
>>> >> To this end I will stop merging patches in to 4.7. Please target the
>>> >> 4.10 tree from here on.
>>> >>
>>> >> Use this thread to discuss concerns and coordinate the work that needs
>>> >> to be done. Bits that I'm aware of
>>> >>
>>> >>  - Eddie is updating userspace to the new FSI APIs
>>> >>  - I need to merge the upstream FSI patchset
>>> >>  - Brednan needs to post the lastest version of the i2c patchset which
>>> >> I will merge
>>> >>  - I need to merge benh's ftgmac100 improvements
>>> >
>>> > We'll need to update the defconfig in the build system. Currently if I
>>> > compile the kernel and generate a new defconfig (make savedefconfig),
>>> > I see the following changes in the 4.10 generated defconfig:
>>> >
>>> > @@ -73,7 +73,6 @@
>>> >  CONFIG_MTD_BLOCK=y
>>> >  CONFIG_MTD_PARTITIONED_MASTER=y
>>> >  CONFIG_MTD_SPI_NOR=y
>>> > -CONFIG_ASPEED_FLASH_SPI=y
>>> >  CONFIG_MTD_UBI=y
>>> >  CONFIG_MTD_UBI_FASTMAP=y
>>> >  CONFIG_MTD_UBI_BLOCK=y
>>> > @@ -116,7 +115,6 @@
>>> >  # CONFIG_SERIO is not set
>>> >  # CONFIG_VT is not set
>>> >  CONFIG_LEGACY_PTY_COUNT=4
>>> > -# CONFIG_DEVKMEM is not set
>>> >  CONFIG_SERIAL_8250=y
>>> >  # CONFIG_SERIAL_8250_DEPRECATED_OPTIONS is not set
>>> >  CONFIG_SERIAL_8250_CONSOLE=y
>>> > @@ -131,10 +129,8 @@
>>> >  CONFIG_I2C=y
>>> >  CONFIG_I2C_CHARDEV=y
>>> >  CONFIG_I2C_ASPEED=y
>>> > -CONFIG_SENSORS_MAX31785=y
>>> >  CONFIG_SENSORS_LM75=y
>>> >  CONFIG_SENSORS_NCT7904=y
>>> > -CONFIG_SENSORS_POWER8_OCC_I2C=y
>>> >  CONFIG_PMBUS=y
>>> >  CONFIG_SENSORS_ADM1275=y
>>> >  CONFIG_SENSORS_UCD9000=y
>>> > @@ -143,21 +139,14 @@
>>> >  # CONFIG_USB_SUPPORT is not set
>>> >  CONFIG_NEW_LEDS=y
>>> >  CONFIG_LEDS_CLASS=y
>>> > -CONFIG_LEDS_GPIO=y
>>> >  CONFIG_LEDS_TRIGGERS=y
>>> >  CONFIG_LEDS_TRIGGER_TIMER=y
>>> >  CONFIG_LEDS_TRIGGER_HEARTBEAT=y
>>> > -CONFIG_LEDS_TRIGGER_GPIO=y
>>> >  CONFIG_RTC_CLASS=y
>>> >  CONFIG_RTC_DRV_DS1307=y
>>> >  CONFIG_RTC_DRV_ASPEED=y
>>> >  CONFIG_MAILBOX=y
>>> > -CONFIG_ASPEED_MBOX=y
>>> >  # CONFIG_IOMMU_SUPPORT is not set
>>> > -CONFIG_FSI=y
>>> > -CONFIG_FSI_MASTER_GPIO=y
>>> > -CONFIG_FSI_SCOM=y
>>> > -CONFIG_FSI_I2C=y
>>> >  CONFIG_FIRMWARE_MEMMAP=y
>>> >  CONFIG_FANOTIFY=y
>>> >  CONFIG_AUTOFS4_FS=y
>>> >
>>> >>
>>> >> Cheers,
>>> >>
>>> >> Joel
>>> >
>>> > I know that CONFIG_ASPEED_FLASH_SPI is now CONFIG_SPI_ASPEED_SMC, but
>>> > I've yet to look at each of the other differences.
>>> >
>>> > cheers
>>> > xo
>>>
>>> OK, I guess it's not that much work to look at those configs. This
>>> seems to work up to the limit of the missing FSI drivers as previously
>>> mentioned:
>>>
>>> ---
>>>
>>> a/meta-openbmc-bsp/meta-aspeed/meta-ast2500/recipes-kernel/linux/linux-obmc/defconfig
>>> +++
>>>
>>> b/meta-openbmc-bsp/meta-aspeed/meta-ast2500/recipes-kernel/linux/linux-obmc/defconfig
>>> @@ -73,7 +73,7 @@ CONFIG_MTD_CMDLINE_PARTS=y
>>>  CONFIG_MTD_BLOCK=y
>>>  CONFIG_MTD_PARTITIONED_MASTER=y
>>>  CONFIG_MTD_SPI_NOR=y
>>> -CONFIG_ASPEED_FLASH_SPI=y
>>> +CONFIG_SPI_ASPEED_SMC=y
>>>  CONFIG_MTD_UBI=y
>>>  CONFIG_MTD_UBI_FASTMAP=y
>>>  CONFIG_MTD_UBI_BLOCK=y
>>> @@ -116,7 +116,6 @@ CONFIG_INPUT_EVDEV=y
>>>  # CONFIG_SERIO is not set
>>>  # CONFIG_VT is not set
>>>  CONFIG_LEGACY_PTY_COUNT=4
>>> -# CONFIG_DEVKMEM is not set
>>>  CONFIG_SERIAL_8250=y
>>>  # CONFIG_SERIAL_8250_DEPRECATED_OPTIONS is not set
>>>  CONFIG_SERIAL_8250_CONSOLE=y
>>> @@ -131,10 +130,11 @@ CONFIG_ASPEED_BT_IPMI_BMC=y
>>>  CONFIG_I2C=y
>>>  CONFIG_I2C_CHARDEV=y
>>>  CONFIG_I2C_ASPEED=y
>>> -CONFIG_SENSORS_MAX31785=y
>>> +CONFIG_GPIOLIB=y
>>> +CONFIG_GPIO_SYSFS=y
>>> +CONFIG_GPIO_ASPEED=y
>>>  CONFIG_SENSORS_LM75=y
>>>  CONFIG_SENSORS_NCT7904=y
>>> -CONFIG_SENSORS_POWER8_OCC_I2C=y
>>>  CONFIG_PMBUS=y
>>>  CONFIG_SENSORS_ADM1275=y
>>>  CONFIG_SENSORS_UCD9000=y
>>> @@ -152,12 +152,8 @@ CONFIG_RTC_CLASS=y
>>>  CONFIG_RTC_DRV_DS1307=y
>>>  CONFIG_RTC_DRV_ASPEED=y
>>>  CONFIG_MAILBOX=y
>>> -CONFIG_ASPEED_MBOX=y
>>> +CONFIG_ASPEED_LPC_MBOX=y
>>>  # CONFIG_IOMMU_SUPPORT is not set
>>> -CONFIG_FSI=y
>>> -CONFIG_FSI_MASTER_GPIO=y
>>> -CONFIG_FSI_SCOM=y
>>> -CONFIG_FSI_I2C=y
>>>  CONFIG_FIRMWARE_MEMMAP=y
>>>  CONFIG_FANOTIFY=y
>>>  CONFIG_AUTOFS4_FS=y
>>>
>>> cheers
>>> xo
>>
>>
>
>
>
>

Re: Switching to 4.10

[Joel Stanley]

On Fri, Apr 28, 2017 at 3:44 AM, Rick Altherr <raltherr@google.com> wrote:
>
>
> On Thu, Apr 27, 2017 at 12:33 AM, Joel Stanley <joel@jms.id.au> wrote:
>>
>> On Thu, Apr 27, 2017 at 3:00 AM, Rick Altherr <raltherr@google.com> wrote:
>> > I saw Joel started https://gerrit.openbmc-project.xyz/#/c/2846/ which
>> > has a
>> > defconfig for ast2400.  If you have one for ast2500, please upload a new
>> > patch set to that change.
>>
>> I have defconfigs in the tree that are ready to go. I will upload them
>> to Gerrit when I get the chance.
>>
>
> I've been holding off on testing some RNG changes because it depends on
> 4.10.  Can you please upload these defconfigs so I can build a full 4.10
> image more easily?

Sure. I will push the 4.10 tree today, and you can use the defconfigs
from there.

Cheers,

Joel

Re: Switching to 4.10

[Matthew Barth]

On 04/27/17 8:24 PM, Joel Stanley wrote:
> On Fri, Apr 28, 2017 at 1:39 AM, Matthew Barth
> <msbarth@linux.vnet.ibm.com> wrote:
>>
>> On 04/27/17 2:33 AM, Joel Stanley wrote:
>>> On Thu, Apr 27, 2017 at 3:00 AM, Rick Altherr <raltherr@google.com> wrote:
>>>> I saw Joel started https://gerrit.openbmc-project.xyz/#/c/2846/ which has
>>>> a
>>>> defconfig for ast2400.  If you have one for ast2500, please upload a new
>>>> patch set to that change.
>>> I have defconfigs in the tree that are ready to go. I will upload them
>>> to Gerrit when I get the chance.
>>>
>>> I've pulled in:
>>>
>>>    - Ben's ftgmac100 patches from the net-next tree
>>>    - Brendan's i2c patches, v7
>>>    - Andrew's GPIO and Pinconf patches from the gpio and pinconf trees
>>>    - Chris Bostic's FSI series, v6
>>>
>>> These are going through a test cycle on the ast2500evb and Plametto
>>> tonight. Once I have an ack for the i2c device tree, I will add that
>>> to the tree and push to github.
>>>
>>> Eddie, can you let me know which version of your OCC hwmon patches you
>>> want included. Same goes for the CFAM I2C patches. If you want to wait
>>> for review then please ping your reviewers and we can see if we can
>>> get the patches reviewed before we include them.
>>>
>>> Brad, can you point me to the SBE FIFO work that you want included. If
>>> you would prefer to wait for review that is okay.
>>>
>>> Matt, I think do you want to merge the MAX31785 hwmon driver as-is for
>>> now?
>> I thought we already have a MAX31785 driver merged, so what is
>> 'as-is'...does that include the update supporting reading the tach feedback
>> from both rotors as described in
>> issue#652(https://github.com/openbmc/openbmc/issues/652)? The commit you
>> referenced in that issue does not have the dual tach feedback support yet.
> No, that driver is not upstream and therefore it is not in the tree.
> It needs to be included in the OpenBMC tree as part of this work.
>
> Thanks for clarifying the issue about dual tach feedback.
>
> Is it useful for you to have the driver merged as-is, or does it need
> to have that feature enabled before you can do any testing?
It would be best to have both tach feedbacks for this. Whatever the 
current driver interface that hwmon is using provides enough 
functionality for reading a single rotor's feedback along with being 
able to set a target speed to the fans currently. Because of this, I'd 
prefer that we focus on getting the dual tach feedback added and merge 
the complete driver then.

Thanks,

Matt
>
> Cheers,
>
> Joel
>
>> At this point in time, it's important we get the MAX31785 driver updated
>> with this support along with the PCA9552 driver presence & led verification
>> of the fan enclosures.
>>

Re: Switching to 4.10

[Brad Bishop]

> On May 1, 2017, at 10:33 AM, Matthew Barth <msbarth@linux.vnet.ibm.com> wrote:
> 
> 
> 
> On 04/27/17 8:24 PM, Joel Stanley wrote:
>> On Fri, Apr 28, 2017 at 1:39 AM, Matthew Barth
>> <msbarth@linux.vnet.ibm.com> wrote:
>>> 
>>> On 04/27/17 2:33 AM, Joel Stanley wrote:
>>>> On Thu, Apr 27, 2017 at 3:00 AM, Rick Altherr <raltherr@google.com> wrote:
>>>>> I saw Joel started https://gerrit.openbmc-project.xyz/#/c/2846/ which has
>>>>> a
>>>>> defconfig for ast2400.  If you have one for ast2500, please upload a new
>>>>> patch set to that change.
>>>> I have defconfigs in the tree that are ready to go. I will upload them
>>>> to Gerrit when I get the chance.
>>>> 
>>>> I've pulled in:
>>>> 
>>>>   - Ben's ftgmac100 patches from the net-next tree
>>>>   - Brendan's i2c patches, v7
>>>>   - Andrew's GPIO and Pinconf patches from the gpio and pinconf trees
>>>>   - Chris Bostic's FSI series, v6
>>>> 
>>>> These are going through a test cycle on the ast2500evb and Plametto
>>>> tonight. Once I have an ack for the i2c device tree, I will add that
>>>> to the tree and push to github.
>>>> 
>>>> Eddie, can you let me know which version of your OCC hwmon patches you
>>>> want included. Same goes for the CFAM I2C patches. If you want to wait
>>>> for review then please ping your reviewers and we can see if we can
>>>> get the patches reviewed before we include them.
>>>> 
>>>> Brad, can you point me to the SBE FIFO work that you want included. If
>>>> you would prefer to wait for review that is okay.
>>>> 
>>>> Matt, I think do you want to merge the MAX31785 hwmon driver as-is for
>>>> now?
>>> I thought we already have a MAX31785 driver merged, so what is
>>> 'as-is'...does that include the update supporting reading the tach feedback
>>> from both rotors as described in
>>> issue#652(https://github.com/openbmc/openbmc/issues/652)? The commit you
>>> referenced in that issue does not have the dual tach feedback support yet.
>> No, that driver is not upstream and therefore it is not in the tree.
>> It needs to be included in the OpenBMC tree as part of this work.
>> 
>> Thanks for clarifying the issue about dual tach feedback.
>> 
>> Is it useful for you to have the driver merged as-is, or does it need
>> to have that feature enabled before you can do any testing?

Sorry Joel, I asked for this to be merged without full understanding.  We are already
testing with a driver that supports a single feedback entry.  So I’m now confused as
to what you would be merging that we don’t already have in the interim before its
reworked to include the extra bytes?

Is it a case of its in our 4.7 branch but not in the 4.10 branch?  In that case please
do merge to 4.10, as we don’t want to go backwards so we can continue to test with what
we have so far.

> It would be best to have both tach feedbacks for this. Whatever the current driver interface that hwmon is using provides enough functionality for reading a single rotor's feedback along with being able to set a target speed to the fans currently. Because of this, I'd prefer that we focus on getting the dual tach feedback added and merge the complete driver then.
> 
> Thanks,
> 
> Matt
>> 
>> Cheers,
>> 
>> Joel
>> 
>>> At this point in time, it's important we get the MAX31785 driver updated
>>> with this support along with the PCA9552 driver presence & led verification
>>> of the fan enclosures.