Received-SPF: pass (google.com: domain of linux-kernel+bounces-104993-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.80.249 as permitted sender) client-ip=147.75.80.249;
From: Thomas Gleixner <tglx@linutronix.de>
To: Adam Dunlap <acdunlap@google.com>, Ingo Molnar <mingo@redhat.com>,
 Borislav Petkov <bp@alien8.de>, Dave Hansen <dave.hansen@linux.intel.com>,
 x86@kernel.org, "H. Peter Anvin" <hpa@zytor.com>, Nathan Chancellor
 <nathan@kernel.org>, Nick Desaulniers <ndesaulniers@google.com>, Bill
 Wendling <morbo@google.com>, Justin Stitt <justinstitt@google.com>, "Peter
 Zijlstra (Intel)" <peterz@infradead.org>, Arjan van de Ven
 <arjan@linux.intel.com>, Wei Liu <wei.liu@kernel.org>,
 linux-kernel@vger.kernel.org, llvm@lists.linux.dev, Jacob Xu
 <jacobhxu@google.com>, Alper Gun <alpergun@google.com>, Kevin Loughlin
 <kevinloughlin@google.com>, Peter Gonda <pgonda@google.com>, Ard
 Biesheuvel <ardb@kernel.org>
Cc: Adam Dunlap <acdunlap@google.com>
Subject: Re: [PATCH v3] x86/asm: Force native_apic_mem_read to use mov
In-Reply-To: <20240206223620.1833276-1-acdunlap@google.com>
References: <20240206223620.1833276-1-acdunlap@google.com>
Date: Sat, 16 Mar 2024 00:44:45 +0100
Message-ID: <87frwrjc0i.ffs@tglx>
Precedence: bulk
MIME-Version: 1.0
Content-Type: text/plain

On Tue, Feb 06 2024 at 14:36, Adam Dunlap wrote:

Can you please use foo() as notation for functions all over the place
including the subject line, which also wants s/mov/the MOV instruction/
and use MOV instead of mov.

> When done from a virtual machine, instructions that touch APIC memory
> must be emulated. By convention, MMIO access are typically performed via
> io.h helpers such as 'readl()' or 'writeq()' to simplify instruction
> emulation/decoding (ex: in KVM hosts and SEV guests) [0].
>
> Currently, native_apic_mem_read does not follow this convention,
> allowing the compiler to emit instructions other than the mov generated
> by readl(). In particular, when compiled with clang and run as a SEV-ES
> or SEV-SNP guest, the compiler would emit a testl instruction which is
> not supported by the SEV-ES emulator, causing a boot failure in that
> environment. It is likely the same problem would happen in a TDX guest
> as that uses the same instruction emulator as SEV-ES.
>
> To make sure all emulators can emulate APIC memory reads via mov, use
> the readl function in native_apic_mem_read. It is expected that any
> emulator would support mov in any addressing mode it is the most generic
> and is what is ususally emitted currently.
>
> The testl instruction is emitted when native_apic_mem_read
> is inlined into __xapic_wait_icr_idle. The emulator comes from
> insn_decode_mmio in arch/x86/lib/insn-eval.c. It would not be worth it

s/It would/It's/

Either it's a fact or not.

> to extend insn_decode_mmio to support more instructions since, in
> theory, the compiler could choose to output nearly any instruction for
> such reads which would bloat the emulator beyond reason.
>
> An alterative to this approach would be to use inline assembly instead
> of the readl helper, as that is what native_apic_mem_write does. I
> consider using readl to be cleaner since it is documented to be a simple
> wrapper and inline assembly is less readable. native_apic_mem_write
> cannot be trivially updated to use writel since it appears to use custom
> asm to workaround for a processor-specific bug.

How is this paragraph relevant?

> [0] https://lore.kernel.org/all/20220405232939.73860-12-kirill.shutemov@linux.intel.com/
>
> Signed-off-by: Adam Dunlap <acdunlap@google.com>
> Tested-by: Kevin Loughlin <kevinloughlin@google.com>

Other than the above nit picks:

Reviewed-by: Thomas Gleixner <tglx@linutronix.de>