Re: [GIT PULL] EFI urgent fixes

From: Ard Biesheuvel <ard.biesheuvel@linaro.org>
To: Ingo Molnar <mingo@kernel.org>
Cc: Matt Fleming <matt@codeblueprint.co.uk>,
	"H. Peter Anvin" <hpa@zytor.com>,
	Thomas Gleixner <tglx@linutronix.de>,
	"linux-efi@vger.kernel.org" <linux-efi@vger.kernel.org>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
	Andy Lutomirski <luto@amacapital.net>,
	Borislav Petkov <bp@alien8.de>,
	Mark Rutland <mark.rutland@arm.com>
Subject: Re: [GIT PULL] EFI urgent fixes
Date: Wed, 18 Feb 2015 16:36:16 +0000	[thread overview]
Message-ID: <D0B4679E-92EB-4A22-8D3E-9D0C27A09ACD@linaro.org> (raw)
In-Reply-To: <20150218134357.GA386@gmail.com>

> On 18 feb. 2015, at 13:43, Ingo Molnar <mingo@kernel.org> wrote:
> 
> 
> * Matt Fleming <matt@codeblueprint.co.uk> wrote:
> 
>> Folks, please pull the following fixes. The revert addresses a
>> regression that Ard hit when booting Qemu and Xen, and the other patch
>> ensures that we don't triple fault when receiving an NMI or MCE during
>> an EFI mixed mode call, which may happen, for instance, when running
>> perf.
>> 
>> The pull is against tip/x86/efi because tip/x86/urgent doesn't contain
>> commit d1a8d66b9177 ("efi/libstub: Call get_memory_map() to obtain map
>> and desc sizes") needed for the revert.
>> 
>> If you'd prefer for me to resend this after the merge window closes,
>> just let me know.
>> 
>> The following changes since commit 3c01b74e818a7a3b2ee9b0d584cca0bc154a031c:
>> 
>>  Merge tag 'efi-next' of git://git.kernel.org/pub/scm/linux/kernel/git/mfleming/efi into x86/efi (2015-01-29 19:16:40 +0100)
>> 
>> are available in the git repository at:
>> 
>> 
>>  git://git.kernel.org/pub/scm/linux/kernel/git/mfleming/efi.git tags/efi-urgent
>> 
>> for you to fetch changes up to 43a9f69692b232d1c64c913a27507eb14a1c47fd:
>> 
>>  Revert "efi/libstub: Call get_memory_map() to obtain map and desc sizes" (2015-02-18 11:38:13 +0000)
>> 
>> ----------------------------------------------------------------
>> * Leave a valid 64-bit IDT installed during runtime EFI mixed mode
>>   calls to avoid triple faults if an NMI/MCE is received.
>> 
>> * Revert Ard's change to the libstub get_memory_map() that went into
>>   the v3.20 merge window because it causes boot regressions on Qemu and
>>   Xen.
>> 
>> ----------------------------------------------------------------
>> Matt Fleming (2):
>>      x86/efi: Avoid triple faults during EFI mixed mode calls
>>      Revert "efi/libstub: Call get_memory_map() to obtain map and desc sizes"
>> 
>> arch/x86/boot/compressed/Makefile              |   1 +
>> arch/x86/boot/compressed/efi_stub_64.S         |  25 ----
>> arch/x86/boot/compressed/efi_thunk_64.S        | 196 +++++++++++++++++++++++++
>> arch/x86/platform/efi/efi_stub_64.S            | 161 --------------------
>> arch/x86/platform/efi/efi_thunk_64.S           | 121 ++++++++++++---
>> drivers/firmware/efi/libstub/efi-stub-helper.c |  16 +-
>> 6 files changed, 307 insertions(+), 213 deletions(-)
>> create mode 100644 arch/x86/boot/compressed/efi_thunk_64.S
> 
> Pulled, thanks Matt!
> 
> Btw., I find the revert a bit sad: in most cases breaking 
> virtualized environments isn't a regression really, it's 
> _them_ who couple to the kernel in an incestuous way that 
> causes the problem: fixes should come from them ...
> 

It potentially breaks physical environments as well, it is just that i spotted it while testing virt ports of Tianocore. And frankly, the patch wasn't entirely correct to begin with

Regards,
Ard.