According to a commit in the 3.4 vendor kernel sources[0] the
ocmemcx_ahb_clk clock "is controlled by RPM and should not be touched by
APPS.".

[0] https://git.codelinaro.org/clo/la/kernel/msm/-/commit/37df5f2d91b4d5768b37fcaacaeea958dd683ebc

And indeed, when using MDSS+GPU+OCMEM on MSM8226 and not using
clk_ignore_unused, when Linux tries to disable the clock the device
crashes and reboots.

And since there's also no evidence of this clock in msm8974 vendor
kernel sources, remove the clock for msm8226 and msm8974.

Fixes: d8b212014e69 ("clk: qcom: Add support for MSM8974's multimedia clock controller (MMCC)")
[message trimmed]

Add more busses found on msm8226 SoC.

Signed-off-by: Luca Weiss <luca@z3ntu.xyz>
---
 arch/arm/boot/dts/qcom/qcom-msm8226.dtsi | 29 +++++++++++++++++++++++++++++
 1 file changed, 29 insertions(+)

diff --git a/arch/arm/boot/dts/qcom/qcom-msm8226.dtsi b/arch/arm/boot/dts/qcom/qcom-msm8226.dtsi
index b6ae4b7936e3..3b6114049e61 100644
--- a/arch/arm/boot/dts/qcom/qcom-msm8226.dtsi
+++ b/arch/arm/boot/dts/qcom/qcom-msm8226.dtsi
@@ -230,6 +230,15 @@ blsp1_uart1: serial@f991d000 {
 			status = "disabled";
 		};
[message trimmed]

On GPIO22 and GPIO23 there is another I2C bus. Add the function for it.

Signed-off-by: Luca Weiss <luca@z3ntu.xyz>
---
 drivers/pinctrl/qcom/pinctrl-msm8226.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/drivers/pinctrl/qcom/pinctrl-msm8226.c b/drivers/pinctrl/qcom/pinctrl-msm8226.c
index 994619840a70..da964eebd313 100644
--- a/drivers/pinctrl/qcom/pinctrl-msm8226.c
+++ b/drivers/pinctrl/qcom/pinctrl-msm8226.c
@@ -336,6 +336,7 @@ enum msm8226_functions {
 	msm_mux_blsp_i2c3,
 	msm_mux_blsp_i2c4,
[message trimmed]

Add the I2C bus and UART interface found on the MSM8226. For the I2C bus
we also first need to add the pinctrl function in the driver.

Signed-off-by: Luca Weiss <luca@z3ntu.xyz>
---
Luca Weiss (2):
      pinctrl: qcom: msm8226: Add blsp_i2c6 function
      ARM: dts: qcom: msm8226: Add blsp1_i2c6 and blsp1_uart2

 arch/arm/boot/dts/qcom/qcom-msm8226.dtsi | 29 +++++++++++++++++++++++++++++
 drivers/pinctrl/qcom/pinctrl-msm8226.c   |  8 ++++++--
 2 files changed, 35 insertions(+), 2 deletions(-)
---
base-commit: 7c2878be573282a9961c359b806ccf70afe1a6b6

On Sonntag, 27. August 2023 19:42:00 CEST Pablo Correa Gómez wrote:
> From: Pablo Correa Gómez <ablocorrea@hotmail.com>
> 
> To use the same convention as in pmaports.
> 
> See https://gitlab.com/postmarketOS/pmaports/-/merge_requests/4283
> 
> Signed-off-by: Pablo Correa Gómez <ablocorrea@hotmail.com>

Reviewed-by: Luca Weiss <luca@z3ntu.xyz>

> ---
>  .../aports/device/testing/device-sony-amami/APKBUILD            | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)

On Samstag, 12. August 2023 17:55:26 CEST Raymond Hackley wrote:

I guess on riscv64 it'll also work for example. armhf maybe even also but 
noone should use that anymore on devices that aren't super old.

I'll let Ollie judge what to include :D

Anyways (assuming it actually works):

Reviewed-by: Luca Weiss <luca@z3ntu.xyz>

> ---
>  README.md | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)

On Sonntag, 6. August 2023 20:43:32 CEST Oliver Smith wrote:
> Remove test_crossdirect_rust, the only test in test_crossdirect.py. The
> behavior that was tested here with all the "mv /usr/bin/rustc" commands
> was specific to version 4 and isn't valid anymore since version 5
> (pmaports MR 4234).
> 
> The previous behavior was that crossdirect always tried to run rustc
> from /native, and if it failed, it would fall back to using qemu. The
> new behavior is that it runs rustc from /native only when it expects it
> to succeed, and runs rustc via qemu otherwise, without attempting
> /native first. As the logic to decide whether to use native rustc or not
> is now done beforehand, it does not do the fallback to qemu if native
> fails.
> 

On Sonntag, 6. August 2023 20:43:31 CEST Oliver Smith wrote:
> Cache the compiler output of rust code with sccache, like we use ccache
> for c code.
> 
> I've considered using sccache to completely replace ccache since it can
> cache output of C/C++ code too. But let's not do it for now since ccache
> doesn't need to run a daemon in the background that needs to be stopped
> when shutting down / zapping. Also it would need more refactoring.

Sounds like a decent idea. Any idea about the performance difference vs ccache 
for C/C++ code? Killing the daemon is already implemented so that's not an 
obstacle anymore.

Also if this is done, "pmbootstrap stats" would need adjustment. Or maybe we

On Sonntag, 6. August 2023 20:43:30 CEST Oliver Smith wrote:

I trust you that this comment is correct ;)

Reviewed-by: Luca Weiss <luca@z3ntu.xyz>

> ---
>  pmb/build/init.py | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/pmb/build/init.py b/pmb/build/init.py
> index f10ae73c..38bfae5f 100644
> --- a/pmb/build/init.py
> +++ b/pmb/build/init.py

On Sonntag, 6. August 2023 20:43:29 CEST Oliver Smith wrote:
> Don't include the /mnt/pmbootstrap directories in the images created
> with "pmbootstrap install".

Don't we have some pmbootstrap/pmaports issues that get resolved here? I swear 
we've had some reports about this.

Anyways:

Reviewed-by: Luca Weiss <luca@z3ntu.xyz>

> ---
>  pmb/chroot/__init__.py    |  2 +-
>  pmb/chroot/mount.py       | 15 +++++++++++++++