Re: [PATCH 12/13] arm64: defconfig: sync with savedefconfig

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 14/12/2023 13:08, Arnd Bergmann wrote:
> On Thu, Dec 14, 2023, at 11:52, Tudor Ambarus wrote:
>> Sync the defconfig with savedefconfig as config options
>> change/move over time.
>>
>> Generated with the following commands:
>> make ARCH=arm64 CROSS_COMPILE=aarch64-linux-gnu- defconfig
>> make ARCH=arm64 CROSS_COMPILE=aarch64-linux-gnu- savedefconfig
>> cp defconfig arch/arm64/configs/defconfig

These are obvious. You cannot do it differently.

>>
>> Signed-off-by: Tudor Ambarus <tudor.ambarus@xxxxxxxxxx>
>> ---
>>  arch/arm64/configs/defconfig | 144 +++++++++++++----------------------
>>  1 file changed, 55 insertions(+), 89 deletions(-)
> 
> I usually ask for defconfig changes to be merged when someone just
> adds a single line per patch, but a 144 line change is clearly too
> big, please split this up.

Anyway this should not go via my tree, because of possible conflicts.
This commit, so the savedefconfig, must be prepared on linux-next, which
should be mentioned in changelog for example. It also is not related to
this patchset.

> 
>> diff --git a/arch/arm64/configs/defconfig b/arch/arm64/configs/defconfig
>> index b60aa1f89343..09fb467303ba 100644
>> --- a/arch/arm64/configs/defconfig
>> +++ b/arch/arm64/configs/defconfig
>> @@ -30,6 +30,8 @@ CONFIG_SCHED_AUTOGROUP=y
>>  CONFIG_BLK_DEV_INITRD=y
>>  CONFIG_KALLSYMS_ALL=y
>>  CONFIG_PROFILING=y
>> +CONFIG_KEXEC_FILE=y
>> +CONFIG_CRASH_DUMP=y
>>  CONFIG_ARCH_ACTIONS=y
>>  CONFIG_ARCH_SUNXI=y
>>  CONFIG_ARCH_ALPINE=y
>> @@ -77,9 +79,6 @@ CONFIG_ARM64_VA_BITS_48=y
>>  CONFIG_SCHED_MC=y
>>  CONFIG_SCHED_SMT=y
>>  CONFIG_NUMA=y
>> -CONFIG_KEXEC=y
>> -CONFIG_KEXEC_FILE=y
>> -CONFIG_CRASH_DUMP=y
>>  CONFIG_XEN=y
>>  CONFIG_COMPAT=y
>>  CONFIG_RANDOMIZE_BASE=y
> 
> These two hunks seem to go together, but it needs an explanation
> why you are removing CONFIG_KEXEC.
> 
>> @@ -119,7 +118,6 @@ CONFIG_KVM=y
>>  CONFIG_JUMP_LABEL=y
>>  CONFIG_MODULES=y
>>  CONFIG_MODULE_UNLOAD=y
>> -CONFIG_IOSCHED_BFQ=y
>>  # CONFIG_CORE_DUMP_DEFAULT_ELF_HEADERS is not set
>>  # CONFIG_COMPAT_BRK is not set
>>  CONFIG_MEMORY_HOTPLUG=y
> 
> No, I definitely want CONFIG_IOSCHED_BFQ=y, it is essential
> for performance on certain classes of machines. It would
> be better to drop the 'imply IOSCHED_BFQ' in two Kconfig
> files.
> 
>> @@ -129,8 +127,6 @@ CONFIG_MEMORY_FAILURE=y
>>  CONFIG_TRANSPARENT_HUGEPAGE=y
>>  CONFIG_NET=y
>>  CONFIG_PACKET=y
>> -CONFIG_UNIX=y
>> -CONFIG_INET=y
>>  CONFIG_IP_MULTICAST=y
>>  CONFIG_IP_PNP=y
>>  CONFIG_IP_PNP_DHCP=y
> 
> These also seem kind of essential for almost any machine,
> I assume you are doing something wrong here.

Yep. I think the folks forget the rule not to remove user-selectable
options :/

> 


...

> 
>>  CONFIG_SLIM_QCOM_NGD_CTRL=m
>> -CONFIG_INTERCONNECT=y
>>  CONFIG_INTERCONNECT_IMX=y
>>  CONFIG_INTERCONNECT_IMX8MM=m
>>  CONFIG_INTERCONNECT_IMX8MN=m
> 
> I think the problem here are some Tegra device drivers that
> incorrectly 'select INTERCONNECT' rather than using the
> 'depends on' that every other interconnect driver has.
> Please fix those instead.

In current setup interconnect is user-selectable, so it must not be removed.

Best regards,
Krzysztof





[Index of Archives]     [Linux GPIO]     [Linux SPI]     [Linux Hardward Monitoring]     [LM Sensors]     [Linux USB Devel]     [Linux Media]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux