diff mbox series

[RESEND,1/2] soc: qcom: rmtfs: allow building the module with COMPILE_TEST=y

Message ID 20241202094903.18388-1-brgl@bgdev.pl
State New
Headers show
Series [RESEND,1/2] soc: qcom: rmtfs: allow building the module with COMPILE_TEST=y | expand

Commit Message

Bartosz Golaszewski Dec. 2, 2024, 9:49 a.m. UTC 
From: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

Make it possible to build the module when COMPILE_TEST is enabled for
better build coverage.

Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
---
 drivers/soc/qcom/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Comments

Bjorn Andersson Dec. 26, 2024, 10:38 p.m. UTC | #1 
On Mon, 02 Dec 2024 10:49:02 +0100, Bartosz Golaszewski wrote:
> Make it possible to build the module when COMPILE_TEST is enabled for
> better build coverage.
> 
> 

Applied, thanks!

[1/2] soc: qcom: rmtfs: allow building the module with COMPILE_TEST=y
      commit: a9020afe40e8752c6db905e895e9002a83255e7b
[2/2] soc: qcom: rmtfs: constify rmtfs_class
      commit: eb9fc0f32f9b4e8b7984c276c1ed6f61d39ef630

Best regards,
Jens Glathe Jan. 9, 2025, 9:55 a.m. UTC | #2 
On 02.12.24 10:49, Bartosz Golaszewski wrote:

> From: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
>
> Make it possible to build the module when COMPILE_TEST is enabled for
> better build coverage.

Hi Bartosz,

no idea why this happens, but I just bisected my way to this patchset
for this very odd error on the HP Omnibook X14:

jglathe@x14-jg:~$ sudo dmesg|grep snps
[    1.009307] qcom-snps-eusb2-hsphy 88e1000.phy: Registered Qcom-eUSB2 phy
[   24.035977] platform fd3000.phy: deferred probe pending:
qcom-snps-eusb2-hsphy: failed to get repeater
[   24.035987] platform fd9000.phy: deferred probe pending:
qcom-snps-eusb2-hsphy: failed to get repeater

This kills USB2 functionality on the type-c connectors.

with best regards

Jens
Bartosz Golaszewski Jan. 9, 2025, 11:13 a.m. UTC | #3 
On Thu, Jan 9, 2025 at 10:56 AM Jens Glathe
<jens.glathe@oldschoolsolutions.biz> wrote:
>
> On 02.12.24 10:49, Bartosz Golaszewski wrote:
>
> > From: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
> >
> > Make it possible to build the module when COMPILE_TEST is enabled for
> > better build coverage.
>
> Hi Bartosz,
>
> no idea why this happens, but I just bisected my way to this patchset
> for this very odd error on the HP Omnibook X14:
>
> jglathe@x14-jg:~$ sudo dmesg|grep snps
> [    1.009307] qcom-snps-eusb2-hsphy 88e1000.phy: Registered Qcom-eUSB2 phy
> [   24.035977] platform fd3000.phy: deferred probe pending:
> qcom-snps-eusb2-hsphy: failed to get repeater
> [   24.035987] platform fd9000.phy: deferred probe pending:
> qcom-snps-eusb2-hsphy: failed to get repeater
>
> This kills USB2 functionality on the type-c connectors.
>
> with best regards
>
> Jens
>

Well, that's weird. Are you 100% sure this commit is the culprit? I
have no idea how this could happen, it's just a Kconfig tweak.

Bartosz
Jens Glathe Jan. 9, 2025, 1:27 p.m. UTC | #4 
On 09.01.25 12:13, Bartosz Golaszewski wrote:
> Well, that's weird. Are you 100% sure this commit is the culprit? I
> have no idea how this could happen, it's just a Kconfig tweak.

That's what I thought, too. I merged qcom/for-next into my branch for
testing and adventure, compared / merged my dt's that were in both, and
then this showed up on the HP X14. On the Snapdragon Dev Kit all 5 eUSB2
repeaters work regardless with the same build. The chip in question is
the SMB2360, a PMIC which also contains the eUSB2 repeater.

jglathe@sdbox2:~/src/linux_ms_dev_kit$ git bisect log
git bisect start
# status: waiting for both good and bad commits
# bad: [a9059ebc60d5b845a95b25650eb24c1df5ed2132] arm64: dts: qcom:
x1e80100: Fix usb_2 controller interrupts
git bisect bad a9059ebc60d5b845a95b25650eb24c1df5ed2132
# status: waiting for good commit(s), bad commit known
# good: [2bab385991536a07d6f4c0970fa665f9626db059] Ubuntu: rebase to
v6.13-rc6
git bisect good 2bab385991536a07d6f4c0970fa665f9626db059
# good: [2cc57105ca61b9fb3744adbf83bbc8b2c0c0d055] johan_defconfig: add
custom config
git bisect good 2cc57105ca61b9fb3744adbf83bbc8b2c0c0d055
# good: [260e63d4eece13d5f908646b8b9e5af44a40d012] arm64: dts: qcom:
msm8996-xiaomi-gemini: Fix LP5562 LED1 reg property
git bisect good 260e63d4eece13d5f908646b8b9e5af44a40d012
# good: [033bf6bd4e245be05553ed1ec774dd78c77922ec] arm64: dts: qcom:
sar2130p: correct sleep clock frequency
git bisect good 033bf6bd4e245be05553ed1ec774dd78c77922ec
# good: [76f8b9a9e0d9fdafa6103ea789865dd6897f3bd6] dt-bindings: clock:
qcom-rpmhcc: Add RPMHCC bindings for QCS615
git bisect good 76f8b9a9e0d9fdafa6103ea789865dd6897f3bd6
# good: [22507018ce51df4f63a5ad468608426bcd799e2b] clk: qcom:
gcc-mdm9607: Fix cmd_rcgr offset for blsp1_uart6 rcg
git bisect good 22507018ce51df4f63a5ad468608426bcd799e2b
# good: [8db2b9a883cbb85a189ad74c3fb3264cc54134b8] soc: qcom:
pmic_glink_altmode: simplify locking with guard()
git bisect good 8db2b9a883cbb85a189ad74c3fb3264cc54134b8
# bad: [17f8cef45b6a390dd2216d28290c21a92c3e3ae1] Ubuntu: Update changelog
git bisect bad 17f8cef45b6a390dd2216d28290c21a92c3e3ae1
# bad: [1c51ac09700c1722329672f6254dca78fbc32101] soc: qcom: rmtfs:
constify rmtfs_class
git bisect bad 1c51ac09700c1722329672f6254dca78fbc32101
# bad: [d6c656205ae9eb32dce89df95faabe898029e668] soc: qcom: rmtfs:
allow building the module with COMPILE_TEST=y
git bisect bad d6c656205ae9eb32dce89df95faabe898029e668
# first bad commit: [d6c656205ae9eb32dce89df95faabe898029e668] soc:
qcom: rmtfs: allow building the module with COMPILE_TEST=y

I have put the tag up on github [1]. It _is_ odd.

Jens

[1]
https://github.com/jglathe/linux_ms_dev_kit/tree/jg/ubuntu-qcom-x1e-6.13-rc6-eUSB2-bad
Konrad Dybcio Jan. 9, 2025, 2:01 p.m. UTC | #5 
On 9.01.2025 2:27 PM, Jens Glathe wrote:
> On 09.01.25 12:13, Bartosz Golaszewski wrote:
>> Well, that's weird. Are you 100% sure this commit is the culprit? I
>> have no idea how this could happen, it's just a Kconfig tweak.
> 
> That's what I thought, too. I merged qcom/for-next into my branch for
> testing and adventure, compared / merged my dt's that were in both, and
> then this showed up on the HP X14. On the Snapdragon Dev Kit all 5 eUSB2
> repeaters work regardless with the same build. The chip in question is
> the SMB2360, a PMIC which also contains the eUSB2 repeater.
> 
> jglathe@sdbox2:~/src/linux_ms_dev_kit$ git bisect log
> git bisect start
> # status: waiting for both good and bad commits
> # bad: [a9059ebc60d5b845a95b25650eb24c1df5ed2132] arm64: dts: qcom:
> x1e80100: Fix usb_2 controller interrupts
> git bisect bad a9059ebc60d5b845a95b25650eb24c1df5ed2132
> # status: waiting for good commit(s), bad commit known
> # good: [2bab385991536a07d6f4c0970fa665f9626db059] Ubuntu: rebase to
> v6.13-rc6
> git bisect good 2bab385991536a07d6f4c0970fa665f9626db059
> # good: [2cc57105ca61b9fb3744adbf83bbc8b2c0c0d055] johan_defconfig: add
> custom config
> git bisect good 2cc57105ca61b9fb3744adbf83bbc8b2c0c0d055
> # good: [260e63d4eece13d5f908646b8b9e5af44a40d012] arm64: dts: qcom:
> msm8996-xiaomi-gemini: Fix LP5562 LED1 reg property
> git bisect good 260e63d4eece13d5f908646b8b9e5af44a40d012
> # good: [033bf6bd4e245be05553ed1ec774dd78c77922ec] arm64: dts: qcom:
> sar2130p: correct sleep clock frequency
> git bisect good 033bf6bd4e245be05553ed1ec774dd78c77922ec
> # good: [76f8b9a9e0d9fdafa6103ea789865dd6897f3bd6] dt-bindings: clock:
> qcom-rpmhcc: Add RPMHCC bindings for QCS615
> git bisect good 76f8b9a9e0d9fdafa6103ea789865dd6897f3bd6
> # good: [22507018ce51df4f63a5ad468608426bcd799e2b] clk: qcom:
> gcc-mdm9607: Fix cmd_rcgr offset for blsp1_uart6 rcg
> git bisect good 22507018ce51df4f63a5ad468608426bcd799e2b
> # good: [8db2b9a883cbb85a189ad74c3fb3264cc54134b8] soc: qcom:
> pmic_glink_altmode: simplify locking with guard()
> git bisect good 8db2b9a883cbb85a189ad74c3fb3264cc54134b8
> # bad: [17f8cef45b6a390dd2216d28290c21a92c3e3ae1] Ubuntu: Update changelog
> git bisect bad 17f8cef45b6a390dd2216d28290c21a92c3e3ae1
> # bad: [1c51ac09700c1722329672f6254dca78fbc32101] soc: qcom: rmtfs:
> constify rmtfs_class
> git bisect bad 1c51ac09700c1722329672f6254dca78fbc32101
> # bad: [d6c656205ae9eb32dce89df95faabe898029e668] soc: qcom: rmtfs:
> allow building the module with COMPILE_TEST=y
> git bisect bad d6c656205ae9eb32dce89df95faabe898029e668
> # first bad commit: [d6c656205ae9eb32dce89df95faabe898029e668] soc:
> qcom: rmtfs: allow building the module with COMPILE_TEST=y
> 
> I have put the tag up on github [1]. It _is_ odd.

I think I've seen this issue before. I'd bet it's just timing.

Konrad
Jens Glathe Jan. 9, 2025, 4:43 p.m. UTC | #6 
Am 09.01.25 um 15:01 schrieb Konrad Dybcio:

> On 9.01.2025 2:27 PM, Jens Glathe wrote:
>> On 09.01.25 12:13, Bartosz Golaszewski wrote:
>>> Well, that's weird. Are you 100% sure this commit is the culprit? I
>>> have no idea how this could happen, it's just a Kconfig tweak.
> I think I've seen this issue before. I'd bet it's just timing.
>
> Konrad

Well there seem to be other (follow-up) symptoms, like laptop not
willing to wake from sleep, higher battery drain. No idea how to tackle
it further, though. drivers/phy/qualcomm/phy-qcom-snps-eusb2.c has no
recent changes.

Jens
diff mbox series

Patch

diff --git a/drivers/soc/qcom/Kconfig b/drivers/soc/qcom/Kconfig
index 74b9121240f89..58e63cf0036ba 100644
--- a/drivers/soc/qcom/Kconfig
+++ b/drivers/soc/qcom/Kconfig
@@ -139,7 +139,7 @@  config QCOM_RAMP_CTRL
 
 config QCOM_RMTFS_MEM
 	tristate "Qualcomm Remote Filesystem memory driver"
-	depends on ARCH_QCOM
+	depends on ARCH_QCOM || COMPILE_TEST
 	select QCOM_SCM
 	help
 	  The Qualcomm remote filesystem memory driver is used for allocating