2019-11-15 09:26:38

by Ruan Shiyang

Subject: [RFC PATCH] x86/acpi: Drop duplicate BOOT table initialization

From: Cao jin <[email protected]>

ACPI BOOT table is initialized in both acpi_boot_table_init &
acpi_boot_init of setup_arch, but its usage is quite late at the end of
start_kernel. It should be safe to drop one of them. Since it is less
related with table init, drop it from there.

Signed-off-by: Cao jin <[email protected]>
Cc: <[email protected]>
Signed-off-by: Shiyang Ruan <[email protected]>
---
arch/x86/kernel/acpi/boot.c | 2 --
1 file changed, 2 deletions(-)

It existed since git repo is built, so it might has its reason? The
patch is not tested since I don't have BOOT table in my firmware.

diff --git a/arch/x86/kernel/acpi/boot.c b/arch/x86/kernel/acpi/boot.c
index 04205ce127a1..ca1c15bb0b48 100644
--- a/arch/x86/kernel/acpi/boot.c
+++ b/arch/x86/kernel/acpi/boot.c
@@ -1558,8 +1558,6 @@ void __init acpi_boot_table_init(void)
return;
}

- acpi_table_parse(ACPI_SIG_BOOT, acpi_parse_sbf);
-
/*
* blacklist may disable ACPI entirely
*/
--
2.21.0

2019-11-29 11:12:13

by Cao jin

Subject: Re: [RFC PATCH] x86/acpi: Drop duplicate BOOT table initialization

Ping?

--
Sincerely,
Cao jin

On 11/15/19 5:24 PM, Shiyang Ruan wrote:
> From: Cao jin <[email protected]>
>
> ACPI BOOT table is initialized in both acpi_boot_table_init &
> acpi_boot_init of setup_arch, but its usage is quite late at the end of
> start_kernel. It should be safe to drop one of them. Since it is less
> related with table init, drop it from there.
>
> Signed-off-by: Cao jin <[email protected]>
> Cc: <[email protected]>
> Signed-off-by: Shiyang Ruan <[email protected]>
> ---
> arch/x86/kernel/acpi/boot.c | 2 --
> 1 file changed, 2 deletions(-)
>
> It existed since git repo is built, so it might has its reason? The
> patch is not tested since I don't have BOOT table in my firmware.
>
> diff --git a/arch/x86/kernel/acpi/boot.c b/arch/x86/kernel/acpi/boot.c
> index 04205ce127a1..ca1c15bb0b48 100644
> --- a/arch/x86/kernel/acpi/boot.c
> +++ b/arch/x86/kernel/acpi/boot.c
> @@ -1558,8 +1558,6 @@ void __init acpi_boot_table_init(void)
> return;
> }
>
> - acpi_table_parse(ACPI_SIG_BOOT, acpi_parse_sbf);
> -
> /*
> * blacklist may disable ACPI entirely
> */
>

2019-12-12 22:15:55

by Rafael J. Wysocki

Subject: Re: [RFC PATCH] x86/acpi: Drop duplicate BOOT table initialization

On Friday, November 15, 2019 10:24:24 AM CET Shiyang Ruan wrote:
> From: Cao jin <[email protected]>
>
> ACPI BOOT table is initialized in both acpi_boot_table_init &
> acpi_boot_init of setup_arch, but its usage is quite late at the end of
> start_kernel. It should be safe to drop one of them. Since it is less
> related with table init, drop it from there.

Well, "It should be safe to drop one of them" is kind of a weak justification.

I need to be convinced that one of them is redundant. At this point I am not.

> Signed-off-by: Cao jin <[email protected]>
> Cc: <[email protected]>
> Signed-off-by: Shiyang Ruan <[email protected]>
> ---
> arch/x86/kernel/acpi/boot.c | 2 --
> 1 file changed, 2 deletions(-)
>
> It existed since git repo is built, so it might has its reason? The
> patch is not tested since I don't have BOOT table in my firmware.
>
> diff --git a/arch/x86/kernel/acpi/boot.c b/arch/x86/kernel/acpi/boot.c
> index 04205ce127a1..ca1c15bb0b48 100644
> --- a/arch/x86/kernel/acpi/boot.c
> +++ b/arch/x86/kernel/acpi/boot.c
> @@ -1558,8 +1558,6 @@ void __init acpi_boot_table_init(void)
> return;
> }
>
> - acpi_table_parse(ACPI_SIG_BOOT, acpi_parse_sbf);
> -
> /*
> * blacklist may disable ACPI entirely
> */
>