cpufreq: qcom-kryo: Fix section annotations
Subsystems (2)
PATCH v3 1/1 cpufreq: qcom-kryo: Fix section annotations
There is currently a warning when building the Kryo cpufreq driver into the kernel image:
WARNING: vmlinux.o(.text+0x8aa424): Section mismatch in reference from the function qcom_cpufreq_kryo_probe() to the function .init.text:qcom_cpufreq_kryo_get_msm_id()
The function qcom_cpufreq_kryo_probe() references
the function __init qcom_cpufreq_kryo_get_msm_id().
This is often because qcom_cpufreq_kryo_probe lacks a __init annotation or the annotation of qcom_cpufreq_kryo_get_msm_id is wrong.
Remove the '__init' annotation from qcom_cpufreq_kryo_get_msm_id so that there is no more mismatch warning.
Additionally, Nick noticed that the remove function was marked as '__init' when it should really be marked as '__exit'.
---
v1 -> v2:
Add '__init' to qcom_cpufreq_kryo_probe instead of removing it from qcom_cpufreq_kryo_get_msm_id.
v2 -> v3:
Go back to removing '__init' from qcom_cpufreq_kryo_get_msm_id instead of adding '__init' to qcom_cpufreq_kryo_probe, which causes another mismatch warning and add the '__init' -> '__exit' conversion noticed by Nick.
The alternative solution is adding '__init' to qcom_cpufreq_kryo_probe, removing the probe definition from qcom_cpufreq_kryo_driver, then changing platform_driver_register to platform_driver_probe in qcom_cpufreq_kryo_init, like [1]. I don't know how that would perform given that I don't have a way to test this device.
[1]: https://lore.kernel.org/lkml/[email protected]/
drivers/cpufreq/qcom-cpufreq-kryo.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/cpufreq/qcom-cpufreq-kryo.c +2 -2
@@ -44,7 +44,7 @@ enum _msm8996_version {struct platform_device *cpufreq_dt_pdev, *kryo_cpufreq_pdev; -static enum _msm8996_version __init qcom_cpufreq_kryo_get_msm_id (void ) +static enum _msm8996_version qcom_cpufreq_kryo_get_msm_id (void ) {size_t len;u32 *msm_id;
@@ -222,7 +222,7 @@ static int __init qcom_cpufreq_kryo_init(void) }module_init (qcom_cpufreq_kryo_init); -static void __init qcom_cpufreq_kryo_exit (void ) +static void __exit qcom_cpufreq_kryo_exit(void) {platform_device_unregister (kryo_cpufreq_pdev);platform_driver_unregister (&qcom_cpufreq_kryo_driver);
Signature · 1 line
2.19.0
Quoted 60 lines · CEST Nathan Chancellor
There is currently a warning when building the Kryo cpufreq driver into the kernel image:
WARNING:
vmlinux.o(.text+0x8aa424): Section mismatch in reference from the functionqcom_cpufreq_kryo_probe()to the function .init.text:qcom_cpufreq_kryo_get_msm_id()
The functionqcom_cpufreq_kryo_probe()references
the function__initqcom_cpufreq_kryo_get_msm_id().
This is often becauseqcom_cpufreq_kryo_probelacks a__initannotation or the annotation ofqcom_cpufreq_kryo_get_msm_idis wrong.Remove the '
__init' annotation fromqcom_cpufreq_kryo_get_msm_idso that there is no more mismatch warning.Additionally, Nick noticed that the remove function was marked as '
__init' when it should really be marked as '__exit'.Fixes 46e2856b8e18 ("cpufreq: Add Kryo CPU scaling driver")Reported-by Nick Desaulniers <[email protected]>Signed-off-by Nathan Chancellor <[email protected]>---
v1 -> v2:
Add '
__init' toqcom_cpufreq_kryo_probeinstead of removing it fromqcom_cpufreq_kryo_get_msm_id.v2 -> v3:
Go back to removing '
__init' fromqcom_cpufreq_kryo_get_msm_idinstead of adding '__init' toqcom_cpufreq_kryo_probe, which causes another mismatch warning and add the '__init' -> '__exit' conversion noticed by Nick.The alternative solution is adding '
__init' toqcom_cpufreq_kryo_probe, removing the probe definition fromqcom_cpufreq_kryo_driver, then changingplatform_driver_registertoplatform_driver_probeinqcom_cpufreq_kryo_init, like [1]. I don't know how that would perform given that I don't have a way to test this device.[1]: https://lore.kernel.org/lkml/[email protected]/
drivers/cpufreq/qcom-cpufreq-kryo.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)drivers/cpufreq/qcom-cpufreq-kryo.c +1
-1@@ -44,7 +44,7 @@ enum _msm8996_version {struct platform_device *cpufreq_dt_pdev, *kryo_cpufreq_pdev; -static enum _msm8996_version __init qcom_cpufreq_kryo_get_msm_id (void ) +static enum _msm8996_version qcom_cpufreq_kryo_get_msm_id (void ) {size_t len;u32 *msm_id;
drivers/cpufreq/qcom-cpufreq-kryo.c static int __init qcom_cpufreq_kryo_init(void) · line 228
Quoted 8 lines
drivers/cpufreq/qcom-cpufreq-kryo.c +1
-1@@ -222,7 +222,7 @@ static int __init qcom_cpufreq_kryo_init(void) }module_init (qcom_cpufreq_kryo_init); -static void __init qcom_cpufreq_kryo_exit (void ) +static void __exit qcom_cpufreq_kryo_exit(void) {platform_device_unregister (kryo_cpufreq_pdev);platform_driver_unregister (&qcom_cpufreq_kryo_driver);
OK, thanks!
I'll queue this up for 4.19-rc6 if there are no objections.
Quoted 11 lines · Nathan Chancellor
There is currently a warning when building the Kryo cpufreq driver into the kernel image:
WARNING:
vmlinux.o(.text+0x8aa424): Section mismatch in reference from the functionqcom_cpufreq_kryo_probe()to the function .init.text:qcom_cpufreq_kryo_get_msm_id()
The functionqcom_cpufreq_kryo_probe()references
the function__initqcom_cpufreq_kryo_get_msm_id().
This is often becauseqcom_cpufreq_kryo_probelacks a__initannotation or the annotation ofqcom_cpufreq_kryo_get_msm_idis wrong.
Quoted 5 lines
Remove the '
__init' annotation fromqcom_cpufreq_kryo_get_msm_idso that there is no more mismatch warning.Additionally, Nick noticed that the remove function was marked as '
__init' when it should really be marked as '__exit'.
I know Rafael asked to merge the two patches into a single one, but I still believe its the best to have two separate patches to fix these two separate issues. For example, picking this patch for an earlier kernel may produce conflicts because the exit code was written at a later point of time and also that the below Fixes tag applies only to the first problem only. And the individual patch for the exit problem is already on the list.
Quoted 3 lines
Fixes 46e2856b8e18 ("cpufreq: Add Kryo CPU scaling driver")Reported-by Nick Desaulniers <[email protected]>Signed-off-by Nathan Chancellor <[email protected]>
Anyway I am fine if Rafael wants to pick it up this way only.
Signature · 1 line
viresh