Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1537220 > unrolled thread

[PATCH 4/4] efi: Skip parsing of EFI_PROPERTIES_TABLE if EFI_MEMORY_ATTRIBUTES_TABLE is detected

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  [PATCH 4/4] efi: Skip parsing of EFI_PROPERTIES_TABLE if EFI_MEMORY_ATTRIBUTES_TABLE is detected Sai Praneeth Prakhya <sai.praneeth.prakhya@intel.com> - 2016-12-06 20:20 +0100
    Re: [PATCH 4/4] efi: Skip parsing of EFI_PROPERTIES_TABLE if  EFI_MEMORY_ATTRIBUTES_TABLE is detected Matt Fleming <matt@codeblueprint.co.uk> - 2016-12-07 14:40 +0100

#1537220 — [PATCH 4/4] efi: Skip parsing of EFI_PROPERTIES_TABLE if EFI_MEMORY_ATTRIBUTES_TABLE is detected

From: Sai Praneeth <sai.praneeth.prakhya@intel.com>

UEFI specification v2.6 recommends not to use
"EFI_PROPERTIES_RUNTIME_MEMORY_PROTECTION_NON_EXECUTABLE_PE_DATA"
attribute of EFI_PROPERTIES_TABLE. Presently, this is the *only* bit
defined in EFI_PROPERTIES_TABLE. This bit implies that EFI Runtime code
and data regions of an executable image are separate and are aligned as
specified in spec. Please refer to "EFI_PROPERTIES_TABLE" in section 4.6
of UEFI specification v2.6 for more information on this table.

UEFI v2.6 introduces EFI_MEMORY_ATTRIBUTES_TABLE and is intended to
replace EFI_PROPERTIES_TABLE. If EFI_MEMORY_ATTRIBUTES_TABLE is found we
skip updating of efi runtime region mappings based on
EFI_PROPERTIES_TABLE, so let's also skip parsing of EFI_PROPERTIES_TABLE
if we find EFI_MEMORY_ATTRIBUTES_TABLE because we are not using this
table anyways. The only caveat here is, if further versions of UEFI spec
adds some more bits (hence some more attributes) to EFI_PROPERTIES_TABLE
then we might need to parse it again, otherwise there is no good in
doing that. We can also expect that the same attributes might be reflected in
EFI_MEMORY_ATTRIBUTES_TABLE and hence saving us from parsing
EFI_PROPERTIES_TABLE again.

Signed-off-by: Sai Praneeth Prakhya <sai.praneeth.prakhya@intel.com>
Cc: Lee, Chun-Yi <jlee@suse.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Ricardo Neri <ricardo.neri@intel.com>
Cc: Matt Fleming <matt@codeblueprint.co.uk>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Ravi Shankar <ravi.v.shankar@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
---
 drivers/firmware/efi/efi.c | 11 +++++++++++
 1 file changed, 11 insertions(+)

diff --git a/drivers/firmware/efi/efi.c b/drivers/firmware/efi/efi.c
index e7d404059b73..e6c6feaa4d78 100644
--- a/drivers/firmware/efi/efi.c
+++ b/drivers/firmware/efi/efi.c
@@ -531,6 +531,17 @@ int __init efi_config_parse_tables(void *config_tables, int count, int sz,
 
 	efi_memattr_init();
 
+	/*
+	 * Since EFI_MEMORY_ATTRIBUTES_TABLE is intended to replace
+	 * EFI_PROPERTIES_TABLE, let's skip parsing of EFI_PROPERTIES_TABLE
+	 * if we find EFI_MEMORY_ATTRIBUTES_TABLE.
+	 * Note: We might need to *re-enable* parsing of EFI_PROPERTIES_TABLE
+	 * if it defines some bits that are not defined in
+	 * EFI_MEMORY_ATTRIBUTES_TABLE.
+	 */
+	if (efi_enabled(EFI_MEM_ATTR))
+		return 0;
+
 	/* Parse the EFI Properties table if it exists */
 	if (efi.properties_table != EFI_INVALID_TABLE_ADDR) {
 		efi_properties_table_t *tbl;
-- 
2.1.4

[toc] | [next] | [standalone]

#1537751 — Re: [PATCH 4/4] efi: Skip parsing of EFI_PROPERTIES_TABLE if EFI_MEMORY_ATTRIBUTES_TABLE is detected

On Tue, 06 Dec, at 11:16:03AM, Sai Praneeth Prakhya wrote:
> From: Sai Praneeth <sai.praneeth.prakhya@intel.com>
> 
> UEFI specification v2.6 recommends not to use
> "EFI_PROPERTIES_RUNTIME_MEMORY_PROTECTION_NON_EXECUTABLE_PE_DATA"
> attribute of EFI_PROPERTIES_TABLE. Presently, this is the *only* bit
> defined in EFI_PROPERTIES_TABLE. This bit implies that EFI Runtime code
> and data regions of an executable image are separate and are aligned as
> specified in spec. Please refer to "EFI_PROPERTIES_TABLE" in section 4.6
> of UEFI specification v2.6 for more information on this table.
> 
> UEFI v2.6 introduces EFI_MEMORY_ATTRIBUTES_TABLE and is intended to
> replace EFI_PROPERTIES_TABLE. If EFI_MEMORY_ATTRIBUTES_TABLE is found we
> skip updating of efi runtime region mappings based on
> EFI_PROPERTIES_TABLE, so let's also skip parsing of EFI_PROPERTIES_TABLE
> if we find EFI_MEMORY_ATTRIBUTES_TABLE because we are not using this
> table anyways. The only caveat here is, if further versions of UEFI spec
> adds some more bits (hence some more attributes) to EFI_PROPERTIES_TABLE
> then we might need to parse it again, otherwise there is no good in
> doing that. We can also expect that the same attributes might be reflected in
> EFI_MEMORY_ATTRIBUTES_TABLE and hence saving us from parsing
> EFI_PROPERTIES_TABLE again.
> 
> Signed-off-by: Sai Praneeth Prakhya <sai.praneeth.prakhya@intel.com>
> Cc: Lee, Chun-Yi <jlee@suse.com>
> Cc: Borislav Petkov <bp@alien8.de>
> Cc: Ricardo Neri <ricardo.neri@intel.com>
> Cc: Matt Fleming <matt@codeblueprint.co.uk>
> Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
> Cc: Ravi Shankar <ravi.v.shankar@intel.com>
> Cc: Fenghua Yu <fenghua.yu@intel.com>
> ---
>  drivers/firmware/efi/efi.c | 11 +++++++++++
>  1 file changed, 11 insertions(+)

I see where you're coming from with this patch, but I think it's
unnecessary. Turning on/off parsing of Table A based on existence of
Table B just seems like extra work.

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web