| Message ID | 1658827473-121156-1-git-send-email-john.garry@huawei.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | arm64: defconfig: Sync some configs with savedefconfig | expand |
On Tue, Jul 26, 2022 at 11:24 AM John Garry <john.garry@huawei.com> wrote: > > Some configs can obviously be removed when sync'ing with savedefconfig, as > follows: > > - config ARCH_BCMBCA is selected by config ARCH_BCM4908 from > arch/arm64/Kconfig.platforms, and the latter is enabled in the defconfig, > so no need to explicitly enable in the defconfig. I double-checked this one and found that this was just added to the defconfig as a transitional entry, with the ARCH_BCM4908 being the one that is the one to go away, as described in 32d3da7d2f44 ("arm64: make ARCH_BCM4908 select ARCH_BCMBCA"), so I think we should leave this one out. > - config SECCOMP was changed to def_bool y in commit 282a181b1a0d (" > seccomp: Move config option SECCOMP to arch/Kconfig"), so no need to > explicitly enable in the defconfig. > > - config MAILBOX is already selected by some drivers enabled in the > defconfig, so no need to explicitly enable. > > - config QRTR was enabled in the defconfig from commit 1bdf91fd2ae82 (" > arm64: defconfig: Enable Qualcomm QRTR"). However until many kernel > versions later in commit 231a136fdf46 ("arm64: defconfig: enable ath11k > driver"), no driver depended on config QRTR - not for building anyway. > In commit 231a136fdf46, config ATH11K_PCI was enabled and this selects > config QRTR, so there is no need to explicitly enable in the defconfig. The rest is fine. If it's ok with you, I'll just apply the other three bits and drop the ARCH_BCMBCA part from the patch and the description. Thanks, Arnd
On 26/07/2022 12:09, Arnd Bergmann wrote: >> config ARCH_BCMBCA is selected by config ARCH_BCM4908 from >> arch/arm64/Kconfig.platforms, and the latter is enabled in the defconfig, >> so no need to explicitly enable in the defconfig. > I double-checked this one and found that this was just added to the defconfig > as a transitional entry, with the ARCH_BCM4908 being the one that is > the one to go away, as described in 32d3da7d2f44 ("arm64: make > ARCH_BCM4908 select ARCH_BCMBCA"), so I think we should leave this > one out. > >> - config SECCOMP was changed to def_bool y in commit 282a181b1a0d (" >> seccomp: Move config option SECCOMP to arch/Kconfig"), so no need to >> explicitly enable in the defconfig. >> >> - config MAILBOX is already selected by some drivers enabled in the >> defconfig, so no need to explicitly enable. >> >> - config QRTR was enabled in the defconfig from commit 1bdf91fd2ae82 (" >> arm64: defconfig: Enable Qualcomm QRTR"). However until many kernel >> versions later in commit 231a136fdf46 ("arm64: defconfig: enable ath11k >> driver"), no driver depended on config QRTR - not for building anyway. >> In commit 231a136fdf46, config ATH11K_PCI was enabled and this selects >> config QRTR, so there is no need to explicitly enable in the defconfig. > The rest is fine. If it's ok with you, I'll just apply the other three bits > and drop the ARCH_BCMBCA part from the patch and the description. ok, fine, thanks!
diff --git a/arch/arm64/configs/defconfig b/arch/arm64/configs/defconfig index 6b17e263ec4a..f34e14752ca7 100644 --- a/arch/arm64/configs/defconfig +++ b/arch/arm64/configs/defconfig @@ -35,7 +35,6 @@ CONFIG_ARCH_ALPINE=y CONFIG_ARCH_APPLE=y CONFIG_ARCH_BCM2835=y CONFIG_ARCH_BCM4908=y -CONFIG_ARCH_BCMBCA=y CONFIG_ARCH_BCM_IPROC=y CONFIG_ARCH_BERLIN=y CONFIG_ARCH_BRCMSTB=y @@ -121,7 +120,6 @@ CONFIG_CRYPTO_AES_ARM64_CE_BLK=y CONFIG_CRYPTO_CHACHA20_NEON=m CONFIG_CRYPTO_AES_ARM64_BS=m CONFIG_JUMP_LABEL=y -CONFIG_SECCOMP=y CONFIG_MODULES=y CONFIG_MODULE_UNLOAD=y # CONFIG_CORE_DUMP_DEFAULT_ELF_HEADERS is not set @@ -175,7 +173,6 @@ CONFIG_NET_CLS_ACT=y CONFIG_NET_ACT_GACT=m CONFIG_NET_ACT_MIRRED=m CONFIG_NET_ACT_GATE=m -CONFIG_QRTR=m CONFIG_QRTR_SMD=m CONFIG_QRTR_TUN=m CONFIG_CAN=m @@ -1104,7 +1101,6 @@ CONFIG_FSL_MC_DPIO=y CONFIG_FSL_RCPM=y CONFIG_MTK_DEVAPC=m CONFIG_MTK_PMIC_WRAP=y -CONFIG_MAILBOX=y CONFIG_QCOM_AOSS_QMP=y CONFIG_QCOM_COMMAND_DB=y CONFIG_QCOM_CPR=y
Some configs can obviously be removed when sync'ing with savedefconfig, as follows: - config ARCH_BCMBCA is selected by config ARCH_BCM4908 from arch/arm64/Kconfig.platforms, and the latter is enabled in the defconfig, so no need to explicitly enable in the defconfig. - config SECCOMP was changed to def_bool y in commit 282a181b1a0d (" seccomp: Move config option SECCOMP to arch/Kconfig"), so no need to explicitly enable in the defconfig. - config MAILBOX is already selected by some drivers enabled in the defconfig, so no need to explicitly enable. - config QRTR was enabled in the defconfig from commit 1bdf91fd2ae82 (" arm64: defconfig: Enable Qualcomm QRTR"). However until many kernel versions later in commit 231a136fdf46 ("arm64: defconfig: enable ath11k driver"), no driver depended on config QRTR - not for building anyway. In commit 231a136fdf46, config ATH11K_PCI was enabled and this selects config QRTR, so there is no need to explicitly enable in the defconfig. Signed-off-by: John Garry <john.garry@huawei.com>