[PATCH] x86/apic/flat64: Add back the early_param("apic", parse_apic)

[PATCH] x86/apic/flat64: Add back the early_param("apic", parse_apic)

[Dexuan Cui]

parse_apic() allows the user to try a different apic driver than the
default one that's automatically chosen. It works for x86_32, but
doesn't work for x86_64 becauase it was removed in 2009 for x86_64 by:
commit 7b38725318f4 ("x86: remove subarchitecture support code"),
whose changelog doesn't explicitly describe the removal for x86_64.

The patch adds back the functionality for x86_64. The intent is mainly
to work around an APIC emulation bug in Hyper-V in the case of kdump:
currently Hyper-V does not honor the disabled state of the local APICs,
and all the IOAPIC-based interrupts may not be delivered to the correct
virtual CPU, if the logical-mode APIC driver is used (the kdump
kernel usually uses the logical-mode APIC driver, since typically
only 1 CPU is active). Luckily the kdump issue can be worked around by
forcing the kdump kernel to use physical mode, before the fix to Hyper-V
becomes widely available.

IMHO the patch is safe because the current default algorithm to choose
the apic driver is unchanged; the patch makes a difference only when
the user specifies the apic= kernel parameter, e.g. "apic=physical flat".

Signed-off-by: Dexuan Cui <decui@microsoft.com>
---
 arch/x86/kernel/apic/apic_flat_64.c | 27 ++++++++++++++++++++++++++-
 1 file changed, 26 insertions(+), 1 deletion(-)

diff --git a/arch/x86/kernel/apic/apic_flat_64.c b/arch/x86/kernel/apic/apic_flat_64.c
index 7862b152a052..efbec63bb01f 100644
--- a/arch/x86/kernel/apic/apic_flat_64.c
+++ b/arch/x86/kernel/apic/apic_flat_64.c
@@ -23,9 +23,34 @@ static struct apic apic_flat;
 struct apic *apic __ro_after_init = &apic_flat;
 EXPORT_SYMBOL_GPL(apic);
 
+static int cmdline_apic __initdata;
+static int __init parse_apic(char *arg)
+{
+	struct apic **drv;
+
+	if (!arg)
+		return -EINVAL;
+
+	for (drv = __apicdrivers; drv < __apicdrivers_end; drv++) {
+		if (!strcmp((*drv)->name, arg)) {
+			apic = *drv;
+			cmdline_apic = 1;
+			return 0;
+		}
+	}
+
+	/* Parsed again by __setup for debug/verbose */
+	return 0;
+}
+early_param("apic", parse_apic);
+
+
 static int flat_acpi_madt_oem_check(char *oem_id, char *oem_table_id)
 {
-	return 1;
+	if (!cmdline_apic)
+		return 1;
+
+	return apic == &apic_flat;
 }
 
 /*
-- 
2.25.1

Re: [PATCH] x86/apic/flat64: Add back the early_param("apic", parse_apic)

[Randy Dunlap]

On 5/28/20 11:37 PM, Dexuan Cui wrote:
> parse_apic() allows the user to try a different apic driver than the
> default one that's automatically chosen. It works for x86_32, but
> doesn't work for x86_64 becauase it was removed in 2009 for x86_64 by:
> commit 7b38725318f4 ("x86: remove subarchitecture support code"),
> whose changelog doesn't explicitly describe the removal for x86_64.
> 
> The patch adds back the functionality for x86_64. The intent is mainly
> to work around an APIC emulation bug in Hyper-V in the case of kdump:
> currently Hyper-V does not honor the disabled state of the local APICs,
> and all the IOAPIC-based interrupts may not be delivered to the correct
> virtual CPU, if the logical-mode APIC driver is used (the kdump
> kernel usually uses the logical-mode APIC driver, since typically
> only 1 CPU is active). Luckily the kdump issue can be worked around by
> forcing the kdump kernel to use physical mode, before the fix to Hyper-V
> becomes widely available.
> 
> IMHO the patch is safe because the current default algorithm to choose
> the apic driver is unchanged; the patch makes a difference only when
> the user specifies the apic= kernel parameter, e.g. "apic=physical flat".
> 
> Signed-off-by: Dexuan Cui <decui@microsoft.com>
> ---
>  arch/x86/kernel/apic/apic_flat_64.c | 27 ++++++++++++++++++++++++++-
>  1 file changed, 26 insertions(+), 1 deletion(-)

Hi,
Looks like you will also need to update
Documentation/admin-guide/kernel-parameters.txt, where it says:

			For X86-32, this can also be used to specify an APIC
			driver name.


> diff --git a/arch/x86/kernel/apic/apic_flat_64.c b/arch/x86/kernel/apic/apic_flat_64.c
> index 7862b152a052..efbec63bb01f 100644
> --- a/arch/x86/kernel/apic/apic_flat_64.c
> +++ b/arch/x86/kernel/apic/apic_flat_64.c
> @@ -23,9 +23,34 @@ static struct apic apic_flat;
>  struct apic *apic __ro_after_init = &apic_flat;
>  EXPORT_SYMBOL_GPL(apic);
>  
> +static int cmdline_apic __initdata;
> +static int __init parse_apic(char *arg)
> +{
> +	struct apic **drv;
> +
> +	if (!arg)
> +		return -EINVAL;
> +
> +	for (drv = __apicdrivers; drv < __apicdrivers_end; drv++) {
> +		if (!strcmp((*drv)->name, arg)) {
> +			apic = *drv;
> +			cmdline_apic = 1;
> +			return 0;
> +		}
> +	}
> +
> +	/* Parsed again by __setup for debug/verbose */
> +	return 0;
> +}
> +early_param("apic", parse_apic);
> +
> +
>  static int flat_acpi_madt_oem_check(char *oem_id, char *oem_table_id)
>  {
> -	return 1;
> +	if (!cmdline_apic)
> +		return 1;
> +
> +	return apic == &apic_flat;
>  }
>  
>  /*
> 


-- 
~Randy

RE: [PATCH] x86/apic/flat64: Add back the early_param("apic", parse_apic)

[Dexuan Cui]

> From: Randy Dunlap <rdunlap@infradead.org>
> Sent: Friday, May 29, 2020 6:33 AM
> Hi,
> Looks like you will also need to update
> Documentation/admin-guide/kernel-parameters.txt, where it says:
> 
> 			For X86-32, this can also be used to specify an APIC
> 			driver name.
> --
> ~Randy

Hi Randy, 
Thanks for you reminder! I just posted a v2 for this with you Cc'd.

Thanks,
-- Dexuan

Re: [PATCH] x86/apic/flat64: Add back the early_param("apic", parse_apic)

[kbuild test robot]

[-- Attachment #1: Type: text/plain, Size: 1962 bytes --]

Hi Dexuan,

Thank you for the patch! Perhaps something to improve:

[auto build test WARNING on tip/x86/core]
[also build test WARNING on tip/auto-latest v5.7-rc7 next-20200529]
[if your patch is applied to the wrong git tree, please drop us a note to help
improve the system. BTW, we also suggest to use '--base' option to specify the
base tree in git format-patch, please see https://stackoverflow.com/a/37406982]

url:    https://github.com/0day-ci/linux/commits/Dexuan-Cui/x86-apic-flat64-Add-back-the-early_param-apic-parse_apic/20200531-065251
base:   https://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git 9cb1fd0efd195590b828b9b865421ad345a4a145
config: x86_64-defconfig (attached as .config)
compiler: clang version 11.0.0 (https://github.com/llvm/llvm-project 2388a096e7865c043e83ece4e26654bd3d1a20d5)
reproduce (this is a W=1 build):
        wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
        chmod +x ~/bin/make.cross
        # install x86_64 cross compiling tool for clang build
        # apt-get install binutils-x86-64-linux-gnu
        # save the attached .config to linux build tree
        COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross ARCH=x86_64 

If you fix the issue, kindly add following tag as appropriate
Reported-by: kbuild test robot <lkp@intel.com>

All warnings (new ones prefixed by >>, old ones prefixed by <<):

>> WARNING: modpost: vmlinux.o(.text+0x55232): Section mismatch in reference from the function flat_acpi_madt_oem_check() to the variable .init.data:disable_timer_pin_1
The function flat_acpi_madt_oem_check() references
the variable __initdata disable_timer_pin_1.
This is often because flat_acpi_madt_oem_check lacks a __initdata
annotation or the annotation of disable_timer_pin_1 is wrong.

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all(a)lists.01.org

[-- Attachment #2: config.gz --]
[-- Type: application/gzip, Size: 29385 bytes --]

Re: [PATCH] x86/apic/flat64: Add back the early_param("apic", parse_apic)

[Dexuan Cui]

[-- Attachment #1: Type: text/plain, Size: 1406 bytes --]

> From: kbuild test robot <lkp@intel.com>
> Sent: Saturday, May 30, 2020 9:27 PM
> [...]
> Hi Dexuan,
> 
> Thank you for the patch! Perhaps something to improve:
> 
> [auto build test WARNING on tip/x86/core]
> [also build test WARNING on tip/auto-latest v5.7-rc7 next-20200529]
> [if your patch is applied to the wrong git tree, please drop us a note to help
> improve the system. BTW, we also suggest to use '--base' option to specify the
> base tree in git format-patch, please see
> [...]
> url: [...]
> base: [...]
> config: x86_64-defconfig (attached as .config)
> compiler: clang version 11.0.0 [...]
> reproduce (this is a W=1 build):
> [...]
> All warnings (new ones prefixed by >>, old ones prefixed by <<):
> 
> >> WARNING: modpost: vmlinux.o(.text+0x55232): Section mismatch in
> reference from the function flat_acpi_madt_oem_check() to the
> variable .init.data:disable_timer_pin_1
> The function flat_acpi_madt_oem_check() references
> the variable __initdata disable_timer_pin_1.
> This is often because flat_acpi_madt_oem_check lacks a __initdata
> annotation or the annotation of disable_timer_pin_1 is wrong.

The warning is not accurate because flat_acpi_madt_oem_check() has nothing
to do with the variable disable_timer_pin_1.

I just posted a v3 patch to fix the warning by adding the __init tag for 
flat_acpi_madt_oem_check().

Thanks,
-- Dexuan

Re: [PATCH] x86/apic/flat64: Add back the early_param("apic", parse_apic)

[kbuild test robot]

[-- Attachment #1: Type: text/plain, Size: 1563 bytes --]

Hi Dexuan,

Thank you for the patch! Perhaps something to improve:

[auto build test WARNING on tip/x86/core]
[also build test WARNING on tip/auto-latest v5.7-rc7 next-20200529]
[if your patch is applied to the wrong git tree, please drop us a note to help
improve the system. BTW, we also suggest to use '--base' option to specify the
base tree in git format-patch, please see https://stackoverflow.com/a/37406982]

url:    https://github.com/0day-ci/linux/commits/Dexuan-Cui/x86-apic-flat64-Add-back-the-early_param-apic-parse_apic/20200531-065251
base:   https://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git 9cb1fd0efd195590b828b9b865421ad345a4a145
config: x86_64-rhel (attached as .config)
compiler: gcc-9 (Debian 9.3.0-13) 9.3.0
reproduce (this is a W=1 build):
        # save the attached .config to linux build tree
        make W=1 ARCH=x86_64 

If you fix the issue, kindly add following tag as appropriate
Reported-by: kbuild test robot <lkp@intel.com>

All warnings (new ones prefixed by >>, old ones prefixed by <<):

>> WARNING: modpost: vmlinux.o(.text+0x6a947): Section mismatch in reference from the function flat_acpi_madt_oem_check() to the variable .init.data:_node_to_pnode
The function flat_acpi_madt_oem_check() references
the variable __initdata _node_to_pnode.
This is often because flat_acpi_madt_oem_check lacks a __initdata
annotation or the annotation of _node_to_pnode is wrong.

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all(a)lists.01.org

[-- Attachment #2: config.gz --]
[-- Type: application/gzip, Size: 44861 bytes --]

Re: [PATCH] x86/apic/flat64: Add back the early_param("apic", parse_apic)

[Dexuan Cui]

[-- Attachment #1: Type: text/plain, Size: 1476 bytes --]

> From: kbuild test robot <lkp@intel.com>
> Sent: Sunday, May 31, 2020 11:37 AM
> 
> Hi Dexuan,
> 
> Thank you for the patch! Perhaps something to improve:
> 
> [auto build test WARNING on tip/x86/core]
> [also build test WARNING on tip/auto-latest v5.7-rc7 next-20200529]
> [if your patch is applied to the wrong git tree, please drop us a note to help
> improve the system. BTW, we also suggest to use '--base' option to specify the
> base tree in git format-patch, please see
> [...]
> url: [...]
> base: [...]
> config: x86_64-rhel (attached as .config)
> compiler: gcc-9 (Debian 9.3.0-13) 9.3.0
> reproduce (this is a W=1 build):
>         # save the attached .config to linux build tree
>         make W=1 ARCH=x86_64
> 
> If you fix the issue, kindly add following tag as appropriate
> Reported-by: kbuild test robot <lkp@intel.com>
> 
> All warnings (new ones prefixed by >>, old ones prefixed by <<):
> 
> >> WARNING: modpost: vmlinux.o(.text+0x6a947): Section mismatch in
> reference from the function flat_acpi_madt_oem_check() to the
> variable .init.data:_node_to_pnode
> The function flat_acpi_madt_oem_check() references
> the variable __initdata _node_to_pnode.
> This is often because flat_acpi_madt_oem_check lacks a __initdata
> annotation or the annotation of _node_to_pnode is wrong.

This is an out-of-date report against v1 of the patch.
I posted v3 about two hours ago, which addressed the warning.

Thanks,
Dexuan