public inbox for linux-crypto@vger.kernel.org
 help / color / mirror / Atom feed
From: Eric Biggers <ebiggers@kernel.org>
To: x86@kernel.org
Cc: linux-kernel@vger.kernel.org, linux-crypto@vger.kernel.org,
	linux-pm@vger.kernel.org, Borislav Petkov <bp@alien8.de>,
	Thomas Gleixner <tglx@linutronix.de>,
	Ayush Jain <Ayush.Jain3@amd.com>,
	Herbert Xu <herbert@gondor.apana.org.au>,
	Ard Biesheuvel <ardb@kernel.org>
Subject: [PATCH 3/3] x86/fpu: Don't support kernel-mode FPU when irqs_disabled()
Date: Fri, 16 May 2025 16:18:58 -0700	[thread overview]
Message-ID: <20250516231858.27899-4-ebiggers@kernel.org> (raw)
In-Reply-To: <20250516231858.27899-1-ebiggers@kernel.org>

From: Eric Biggers <ebiggers@google.com>

Make irq_fpu_usable() return false when irqs_disabled().  That makes the
irqs_disabled() checks in kernel_fpu_begin_mask() and kernel_fpu_end()
unnecessary, so also remove those.

Rationale:

- There's no known use case for kernel-mode FPU when irqs_disabled().
  arm64 and riscv already disallow kernel-mode FPU when irqs_disabled().
  __save_processor_state() previously did expect kernel_fpu_begin() and
  kernel_fpu_end() to work when irqs_disabled(), but this was a
  different use case and not actual kernel-mode FPU use.

- This is more efficient, since one call to irqs_disabled() replaces two
  irqs_disabled() and one in_hardirq().

- This fixes irq_fpu_usable() to correctly return false during CPU
  initialization.  Incorrectly returning true caused the SHA-256 library
  code, which is called when loading AMD microcode, to take a
  SIMD-optimized code path too early, causing a crash.  By correctly
  returning false from irq_fpu_usable(), the generic SHA-256 code
  correctly gets used instead.  (Note: SIMD-optimized SHA-256 doesn't
  get enabled until subsys_initcall, but CPU hotplug can happen later.)

Fixes: 11d7956d526f ("crypto: x86/sha256 - implement library instead of shash")
Reported-by: Ayush Jain <Ayush.Jain3@amd.com>
Closes: https://lore.kernel.org/r/20250516112217.GBaCcf6Yoc6LkIIryP@fat_crate.local
Signed-off-by: Eric Biggers <ebiggers@google.com>
---
 arch/x86/kernel/fpu/core.c | 49 ++++++++++++++------------------------
 1 file changed, 18 insertions(+), 31 deletions(-)

diff --git a/arch/x86/kernel/fpu/core.c b/arch/x86/kernel/fpu/core.c
index 476393b1d5e8f..9b3c5e17f86cd 100644
--- a/arch/x86/kernel/fpu/core.c
+++ b/arch/x86/kernel/fpu/core.c
@@ -66,42 +66,31 @@ struct fpu *x86_task_fpu(struct task_struct *task)
  * Can we use the FPU in kernel mode with the
  * whole "kernel_fpu_begin/end()" sequence?
  */
 bool irq_fpu_usable(void)
 {
-	if (WARN_ON_ONCE(in_nmi()))
-		return false;
-
 	/*
-	 * In kernel FPU usage already active?  This detects any explicitly
-	 * nested usage in task or softirq context, which is unsupported.  It
-	 * also detects attempted usage in a hardirq that has interrupted a
-	 * kernel-mode FPU section.
+	 * To ensure that (non-explicitly-nested) kernel-mode FPU is always
+	 * usable in task and softirq contexts, kernel_fpu_begin() disables
+	 * preemption and softirqs, and kernel_fpu_end() re-enables them.  That
+	 * is not compatible with hardirqs being disabled (including hardirq
+	 * context), or with NMI context.  Support for kernel-mode FPU is not
+	 * needed in those contexts anyway.  Return false in those contexts.
+	 *
+	 * Returning false when irqs_disabled() also eliminates the need to
+	 * explicitly check whether the FPU has been initialized yet during CPU
+	 * initialization.  Before then, hardirqs are still disabled.
 	 */
+	if (irqs_disabled() || WARN_ON_ONCE(in_nmi()))
+		return false;
+
+	/* Catch any explicitly nested usage, which should never happen. */
 	if (this_cpu_read(in_kernel_fpu)) {
-		WARN_ON_FPU(!in_hardirq());
+		WARN_ON_FPU(1);
 		return false;
 	}
-
-	/*
-	 * When not in NMI or hard interrupt context, FPU can be used in:
-	 *
-	 * - Task context except from within fpregs_lock()'ed critical
-	 *   regions.
-	 *
-	 * - Soft interrupt processing context which cannot happen
-	 *   while in a fpregs_lock()'ed critical region.
-	 */
-	if (!in_hardirq())
-		return true;
-
-	/*
-	 * In hard interrupt context it's safe when soft interrupts
-	 * are enabled, which means the interrupt did not hit in
-	 * a fpregs_lock()'ed critical region.
-	 */
-	return !softirq_count();
+	return true;
 }
 EXPORT_SYMBOL(irq_fpu_usable);
 
 /*
  * Track AVX512 state use because it is known to slow the max clock
@@ -443,12 +432,11 @@ static __always_inline void __fpu_save_state(void)
 	__cpu_invalidate_fpregs_state();
 }
 
 void kernel_fpu_begin_mask(unsigned int kfpu_mask)
 {
-	if (!irqs_disabled())
-		fpregs_lock();
+	fpregs_lock();
 
 	WARN_ON_FPU(!irq_fpu_usable());
 	WARN_ON_FPU(this_cpu_read(in_kernel_fpu));
 
 	this_cpu_write(in_kernel_fpu, true);
@@ -467,12 +455,11 @@ EXPORT_SYMBOL_GPL(kernel_fpu_begin_mask);
 void kernel_fpu_end(void)
 {
 	WARN_ON_FPU(!this_cpu_read(in_kernel_fpu));
 
 	this_cpu_write(in_kernel_fpu, false);
-	if (!irqs_disabled())
-		fpregs_unlock();
+	fpregs_unlock();
 }
 EXPORT_SYMBOL_GPL(kernel_fpu_end);
 
 #ifdef CONFIG_PM_SLEEP
 /*
-- 
2.49.0


  parent reply	other threads:[~2025-05-16 23:21 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-05-16 23:18 [PATCH 0/3] x86: Don't support kernel-mode FPU with hardirqs disabled Eric Biggers
2025-05-16 23:18 ` [PATCH 1/3] x86/fpu: Add fpu_save_state() for __save_processor_state() Eric Biggers
2025-05-16 23:18 ` [PATCH 2/3] x86/pm: Use fpu_save_state() in __save_processor_state() Eric Biggers
2025-05-16 23:18 ` Eric Biggers [this message]
2025-05-17  7:09   ` [PATCH 3/3] x86/fpu: Don't support kernel-mode FPU when irqs_disabled() Ingo Molnar
2025-05-17 18:39     ` Eric Biggers
2025-05-18  6:34       ` Ingo Molnar
2025-05-18 13:18         ` Ard Biesheuvel
2025-05-18 20:01           ` Eric Biggers
2025-05-19  8:05             ` Ingo Molnar
2025-05-19  9:49               ` Ard Biesheuvel
2025-05-19 12:57                 ` Ingo Molnar
2025-05-19 13:50                   ` Ard Biesheuvel
2025-05-20  7:42                     ` Ingo Molnar
2025-05-17  1:30 ` [PATCH 0/3] x86: Don't support kernel-mode FPU with hardirqs disabled Eric Biggers

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20250516231858.27899-4-ebiggers@kernel.org \
    --to=ebiggers@kernel.org \
    --cc=Ayush.Jain3@amd.com \
    --cc=ardb@kernel.org \
    --cc=bp@alien8.de \
    --cc=herbert@gondor.apana.org.au \
    --cc=linux-crypto@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-pm@vger.kernel.org \
    --cc=tglx@linutronix.de \
    --cc=x86@kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox