Let's enable ACPI_HMAT, ACPI_HOTPLUG_MEMORY, MEMORY_HOTPLUG
and MEMORY_HOTREMOVE for more test coverage, also there are
useful for heterogeneous memory scene.
Signed-off-by: Kefeng Wang <[email protected]>
---
v2: add ACPI_HMAT and update changelog too.
arch/arm64/configs/defconfig | 4 ++++
1 file changed, 4 insertions(+)
diff --git a/arch/arm64/configs/defconfig b/arch/arm64/configs/defconfig
index 7d1105343bc2..78add6eefa2b 100644
--- a/arch/arm64/configs/defconfig
+++ b/arch/arm64/configs/defconfig
@@ -101,6 +101,8 @@ CONFIG_ARM_TEGRA186_CPUFREQ=y
CONFIG_ARM_MEDIATEK_CPUFREQ=y
CONFIG_QORIQ_CPUFREQ=y
CONFIG_ACPI=y
+CONFIG_ACPI_HOTPLUG_MEMORY=y
+CONFIG_ACPI_HMAT=y
CONFIG_ACPI_APEI=y
CONFIG_ACPI_APEI_GHES=y
CONFIG_ACPI_APEI_PCIEAER=y
@@ -125,6 +127,8 @@ CONFIG_SECCOMP=y
CONFIG_MODULES=y
CONFIG_MODULE_UNLOAD=y
# CONFIG_CORE_DUMP_DEFAULT_ELF_HEADERS is not set
+CONFIG_MEMORY_HOTPLUG=y
+CONFIG_MEMORY_HOTREMOVE=y
CONFIG_KSM=y
CONFIG_MEMORY_FAILURE=y
CONFIG_TRANSPARENT_HUGEPAGE=y
--
2.35.3
On 2022/6/29 17:35, Kefeng Wang wrote:
> Let's enable ACPI_HMAT, ACPI_HOTPLUG_MEMORY, MEMORY_HOTPLUG
> and MEMORY_HOTREMOVE for more test coverage, also there are
> useful for heterogeneous memory scene.
Hi Catalin and Will，is the defconfig patch picked up from you directly，
the changelog shows most of them merged by SoC maintainers, but this
one is some general feature, I don't know who will take it, thanks.
> Signed-off-by: Kefeng Wang <[email protected]>
> ---
> v2: add ACPI_HMAT and update changelog too.
> arch/arm64/configs/defconfig | 4 ++++
> 1 file changed, 4 insertions(+)
>
> diff --git a/arch/arm64/configs/defconfig b/arch/arm64/configs/defconfig
> index 7d1105343bc2..78add6eefa2b 100644
> --- a/arch/arm64/configs/defconfig
> +++ b/arch/arm64/configs/defconfig
> @@ -101,6 +101,8 @@ CONFIG_ARM_TEGRA186_CPUFREQ=y
> CONFIG_ARM_MEDIATEK_CPUFREQ=y
> CONFIG_QORIQ_CPUFREQ=y
> CONFIG_ACPI=y
> +CONFIG_ACPI_HOTPLUG_MEMORY=y
> +CONFIG_ACPI_HMAT=y
> CONFIG_ACPI_APEI=y
> CONFIG_ACPI_APEI_GHES=y
> CONFIG_ACPI_APEI_PCIEAER=y
> @@ -125,6 +127,8 @@ CONFIG_SECCOMP=y
> CONFIG_MODULES=y
> CONFIG_MODULE_UNLOAD=y
> # CONFIG_CORE_DUMP_DEFAULT_ELF_HEADERS is not set
> +CONFIG_MEMORY_HOTPLUG=y
> +CONFIG_MEMORY_HOTREMOVE=y
> CONFIG_KSM=y
> CONFIG_MEMORY_FAILURE=y
> CONFIG_TRANSPARENT_HUGEPAGE=y
On 6/29/22 15:05, Kefeng Wang wrote:
> Let's enable ACPI_HMAT, ACPI_HOTPLUG_MEMORY, MEMORY_HOTPLUG
> and MEMORY_HOTREMOVE for more test coverage, also there are
> useful for heterogeneous memory scene.
>
> Signed-off-by: Kefeng Wang <[email protected]>
> ---
> v2: add ACPI_HMAT and update changelog too.
In general, would like to see MEMORY_HOTPLUG/HOTREMOVE be selected
here for better memory hotplug test coverage. But not sure whether
there had been a rationale for not enabling this on defconfig.
> arch/arm64/configs/defconfig | 4 ++++
> 1 file changed, 4 insertions(+)
>
> diff --git a/arch/arm64/configs/defconfig b/arch/arm64/configs/defconfig
> index 7d1105343bc2..78add6eefa2b 100644
> --- a/arch/arm64/configs/defconfig
> +++ b/arch/arm64/configs/defconfig
> @@ -101,6 +101,8 @@ CONFIG_ARM_TEGRA186_CPUFREQ=y
> CONFIG_ARM_MEDIATEK_CPUFREQ=y
> CONFIG_QORIQ_CPUFREQ=y
> CONFIG_ACPI=y
> +CONFIG_ACPI_HOTPLUG_MEMORY=y
> +CONFIG_ACPI_HMAT=y
> CONFIG_ACPI_APEI=y
> CONFIG_ACPI_APEI_GHES=y
> CONFIG_ACPI_APEI_PCIEAER=y
> @@ -125,6 +127,8 @@ CONFIG_SECCOMP=y
> CONFIG_MODULES=y
> CONFIG_MODULE_UNLOAD=y
> # CONFIG_CORE_DUMP_DEFAULT_ELF_HEADERS is not set
> +CONFIG_MEMORY_HOTPLUG=y
> +CONFIG_MEMORY_HOTREMOVE=y
> CONFIG_KSM=y
> CONFIG_MEMORY_FAILURE=y
> CONFIG_TRANSPARENT_HUGEPAGE=y
On Thu, Jun 30, 2022 at 12:40:44PM +0530, Anshuman Khandual wrote:
>
>
> On 6/29/22 15:05, Kefeng Wang wrote:
> > Let's enable ACPI_HMAT, ACPI_HOTPLUG_MEMORY, MEMORY_HOTPLUG
> > and MEMORY_HOTREMOVE for more test coverage, also there are
> > useful for heterogeneous memory scene.
> >
> > Signed-off-by: Kefeng Wang <[email protected]>
> > ---
> > v2: add ACPI_HMAT and update changelog too.
>
> In general, would like to see MEMORY_HOTPLUG/HOTREMOVE be selected
> here for better memory hotplug test coverage. But not sure whether
> there had been a rationale for not enabling this on defconfig.
Probably we didn't support the first defconfig was added and forgot
about it.
--
Catalin
On Thu, Jun 30, 2022 at 09:26:17AM +0800, Kefeng Wang wrote:
>
> On 2022/6/29 17:35, Kefeng Wang wrote:
> > Let's enable ACPI_HMAT, ACPI_HOTPLUG_MEMORY, MEMORY_HOTPLUG
> > and MEMORY_HOTREMOVE for more test coverage, also there are
> > useful for heterogeneous memory scene.
>
> Hi Catalin and Will，is the defconfig patch picked up from you directly，
> the changelog shows most of them merged by SoC maintainers, but this
> one is some general feature, I don't know who will take it, thanks.
In general we leave the defconfig patches to the SoC team to avoid
conflicts as they have a lot more changes. Cc'ing Arnd if he wants to
pick it up, otherwise it can go through the arm64 tree.
Reviewed-by: Catalin Marinas <[email protected]>
On 2022/6/30 18:53, Catalin Marinas wrote:
> On Thu, Jun 30, 2022 at 09:26:17AM +0800, Kefeng Wang wrote:
>> On 2022/6/29 17:35, Kefeng Wang wrote:
>>> Let's enable ACPI_HMAT, ACPI_HOTPLUG_MEMORY, MEMORY_HOTPLUG
>>> and MEMORY_HOTREMOVE for more test coverage, also there are
>>> useful for heterogeneous memory scene.
>> Hi Catalin and Will，is the defconfig patch picked up from you directly，
>> the changelog shows most of them merged by SoC maintainers, but this
>> one is some general feature, I don't know who will take it, thanks.
> In general we leave the defconfig patches to the SoC team to avoid
> conflicts as they have a lot more changes. Cc'ing Arnd if he wants to
> pick it up, otherwise it can go through the arm64 tree.
Get it, so Arnd, could you pick it up, thanks.
>
> Reviewed-by: Catalin Marinas <[email protected]>
Thanks.
> .
On 2022/6/30 18:53, Catalin Marinas wrote:
> On Thu, Jun 30, 2022 at 09:26:17AM +0800, Kefeng Wang wrote:
>> On 2022/6/29 17:35, Kefeng Wang wrote:
>>> Let's enable ACPI_HMAT, ACPI_HOTPLUG_MEMORY, MEMORY_HOTPLUG
>>> and MEMORY_HOTREMOVE for more test coverage, also there are
>>> useful for heterogeneous memory scene.
>> Hi Catalin and Will，is the defconfig patch picked up from you directly，
>> the changelog shows most of them merged by SoC maintainers, but this
>> one is some general feature, I don't know who will take it, thanks.
> In general we leave the defconfig patches to the SoC team to avoid
> conflicts as they have a lot more changes. Cc'ing Arnd if he wants to
> pick it up, otherwise it can go through the arm64 tree.
>
> Reviewed-by: Catalin Marinas <[email protected]>
> .
Hi maintainers, look this patch is missed, could anyone pick it up,
thanks.
On Thu, Aug 25, 2022 at 09:09:17PM +0800, Kefeng Wang wrote:
> On 2022/6/30 18:53, Catalin Marinas wrote:
> > On Thu, Jun 30, 2022 at 09:26:17AM +0800, Kefeng Wang wrote:
> > > On 2022/6/29 17:35, Kefeng Wang wrote:
> > > > Let's enable ACPI_HMAT, ACPI_HOTPLUG_MEMORY, MEMORY_HOTPLUG
> > > > and MEMORY_HOTREMOVE for more test coverage, also there are
> > > > useful for heterogeneous memory scene.
> > > Hi Catalin and Will，is the defconfig patch picked up from you directly，
> > > the changelog shows most of them merged by SoC maintainers, but this
> > > one is some general feature, I don't know who will take it, thanks.
> > In general we leave the defconfig patches to the SoC team to avoid
> > conflicts as they have a lot more changes. Cc'ing Arnd if he wants to
> > pick it up, otherwise it can go through the arm64 tree.
> >
> > Reviewed-by: Catalin Marinas <[email protected]>
>
> Hi maintainers, look this patch is missed, could anyone pick it up,
I added it to my list of patches for 6.1. I'll queue it around rc6
unless Arnd takes it before (and of course, if it won't conflict to the
other SoC changes to defconfig).
--
Catalin
On 2022/9/1 18:38, Catalin Marinas wrote:
> On Thu, Aug 25, 2022 at 09:09:17PM +0800, Kefeng Wang wrote:
>> On 2022/6/30 18:53, Catalin Marinas wrote:
>>> On Thu, Jun 30, 2022 at 09:26:17AM +0800, Kefeng Wang wrote:
>>>> On 2022/6/29 17:35, Kefeng Wang wrote:
>>>>> Let's enable ACPI_HMAT, ACPI_HOTPLUG_MEMORY, MEMORY_HOTPLUG
>>>>> and MEMORY_HOTREMOVE for more test coverage, also there are
>>>>> useful for heterogeneous memory scene.
>>>> Hi Catalin and Will，is the defconfig patch picked up from you directly，
>>>> the changelog shows most of them merged by SoC maintainers, but this
>>>> one is some general feature, I don't know who will take it, thanks.
>>> In general we leave the defconfig patches to the SoC team to avoid
>>> conflicts as they have a lot more changes. Cc'ing Arnd if he wants to
>>> pick it up, otherwise it can go through the arm64 tree.
>>>
>>> Reviewed-by: Catalin Marinas <[email protected]>
>> Hi maintainers, look this patch is missed, could anyone pick it up,
> I added it to my list of patches for 6.1. I'll queue it around rc6
> unless Arnd takes it before (and of course, if it won't conflict to the
> other SoC changes to defconfig).
Thanks for letting me know that. if conflict, I could rebase and repost.
On Wed, 29 Jun 2022 17:35:24 +0800, Kefeng Wang wrote:
> Let's enable ACPI_HMAT, ACPI_HOTPLUG_MEMORY, MEMORY_HOTPLUG
> and MEMORY_HOTREMOVE for more test coverage, also there are
> useful for heterogeneous memory scene.
>
>
Applied to arm64 (for-next/defconfig), thanks!
[1/1] arm64: defconfig: Enable memory hotplug and hotremove config
https://git.kernel.org/arm64/c/31dbadcc2828
--
Catalin