arm64: neon: Remove support for nested or hardirq kernel-mode NEON

Support for kernel-mode NEON to be nested and/or used in hardirq
context adds significant complexity, and the benefits may be
marginal.  In practice, kernel-mode NEON is not used in hardirq
context, and is rarely used in softirq context (by certain mac80211
drivers).

This patch implements an arm64 may_use_simd() function to allow
clients to check whether kernel-mode NEON is usable in the current
context, and simplifies kernel_neon_{begin,end}() to handle only
saving of the task FPSIMD state (if any).  Without nesting, there
is no other state to save.

The partial fpsimd save/restore functions become redundant as a
result of these changes, so they are removed too.

The save/restore model is changed to operate directly on
task_struct without additional percpu storage.  This simplifies the
code and saves a bit of memory, but means that softirqs must now be
disabled when manipulating the task fpsimd state from task context:
correspondingly, preempt_{en,dis}sable() calls are upgraded to
local_bh_{en,dis}able() as appropriate.  fpsimd_thread_switch()
already runs with hardirqs disabled and so is already protected
from softirqs.

These changes should make it easier to support kernel-mode NEON in
the presence of the Scalable Vector extension in the future.

Signed-off-by: Dave Martin <Dave.Martin@arm.com>
Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>

4 files changed, 33 insertions, 74 deletions

diff --git a/arch/arm64/include/asm/fpsimd.h b/arch/arm64/include/asm/fpsimd.h
index 5155f21e15e3..410c48163c6a 100644
--- a/arch/arm64/include/asm/fpsimd.h
+++ b/arch/arm64/include/asm/fpsimd.h
@@ -41,16 +41,6 @@ struct fpsimd_state {
 	unsigned int cpu;
 };
 
-/*
- * Struct for stacking the bottom 'n' FP/SIMD registers.
- */
-struct fpsimd_partial_state {
-	u32		fpsr;
-	u32		fpcr;
-	u32		num_regs;
-	__uint128_t	vregs[32];
-};
-
 
 #if defined(__KERNEL__) && defined(CONFIG_COMPAT)
 /* Masks for extracting the FPSR and FPCR from the FPSCR */
@@ -77,10 +67,6 @@ extern void fpsimd_update_current_state(struct fpsimd_state *state);
 
 extern void fpsimd_flush_task_state(struct task_struct *target);
 
-extern void fpsimd_save_partial_state(struct fpsimd_partial_state *state,
-				      u32 num_regs);
-extern void fpsimd_load_partial_state(struct fpsimd_partial_state *state);
-
 /* For use by EFI runtime services calls only */
 extern void __efi_fpsimd_begin(void);
 extern void __efi_fpsimd_end(void);
diff --git a/arch/arm64/include/asm/fpsimdmacros.h b/arch/arm64/include/asm/fpsimdmacros.h
index a2daf1293028..0f5fdd388b0d 100644
--- a/arch/arm64/include/asm/fpsimdmacros.h
+++ b/arch/arm64/include/asm/fpsimdmacros.h
@@ -75,59 +75,3 @@
 	ldr	w\tmpnr, [\state, #16 * 2 + 4]
 	fpsimd_restore_fpcr x\tmpnr, \state
 .endm
-
-.macro fpsimd_save_partial state, numnr, tmpnr1, tmpnr2
-	mrs	x\tmpnr1, fpsr
-	str	w\numnr, [\state, #8]
-	mrs	x\tmpnr2, fpcr
-	stp	w\tmpnr1, w\tmpnr2, [\state]
-	adr	x\tmpnr1, 0f
-	add	\state, \state, x\numnr, lsl #4
-	sub	x\tmpnr1, x\tmpnr1, x\numnr, lsl #1
-	br	x\tmpnr1
-	stp	q30, q31, [\state, #-16 * 30 - 16]
-	stp	q28, q29, [\state, #-16 * 28 - 16]
-	stp	q26, q27, [\state, #-16 * 26 - 16]
-	stp	q24, q25, [\state, #-16 * 24 - 16]
-	stp	q22, q23, [\state, #-16 * 22 - 16]
-	stp	q20, q21, [\state, #-16 * 20 - 16]
-	stp	q18, q19, [\state, #-16 * 18 - 16]
-	stp	q16, q17, [\state, #-16 * 16 - 16]
-	stp	q14, q15, [\state, #-16 * 14 - 16]
-	stp	q12, q13, [\state, #-16 * 12 - 16]
-	stp	q10, q11, [\state, #-16 * 10 - 16]
-	stp	q8, q9, [\state, #-16 * 8 - 16]
-	stp	q6, q7, [\state, #-16 * 6 - 16]
-	stp	q4, q5, [\state, #-16 * 4 - 16]
-	stp	q2, q3, [\state, #-16 * 2 - 16]
-	stp	q0, q1, [\state, #-16 * 0 - 16]
-0:
-.endm
-
-.macro fpsimd_restore_partial state, tmpnr1, tmpnr2
-	ldp	w\tmpnr1, w\tmpnr2, [\state]
-	msr	fpsr, x\tmpnr1
-	fpsimd_restore_fpcr x\tmpnr2, x\tmpnr1
-	adr	x\tmpnr1, 0f
-	ldr	w\tmpnr2, [\state, #8]
-	add	\state, \state, x\tmpnr2, lsl #4
-	sub	x\tmpnr1, x\tmpnr1, x\tmpnr2, lsl #1
-	br	x\tmpnr1
-	ldp	q30, q31, [\state, #-16 * 30 - 16]
-	ldp	q28, q29, [\state, #-16 * 28 - 16]
-	ldp	q26, q27, [\state, #-16 * 26 - 16]
-	ldp	q24, q25, [\state, #-16 * 24 - 16]
-	ldp	q22, q23, [\state, #-16 * 22 - 16]
-	ldp	q20, q21, [\state, #-16 * 20 - 16]
-	ldp	q18, q19, [\state, #-16 * 18 - 16]
-	ldp	q16, q17, [\state, #-16 * 16 - 16]
-	ldp	q14, q15, [\state, #-16 * 14 - 16]
-	ldp	q12, q13, [\state, #-16 * 12 - 16]
-	ldp	q10, q11, [\state, #-16 * 10 - 16]
-	ldp	q8, q9, [\state, #-16 * 8 - 16]
-	ldp	q6, q7, [\state, #-16 * 6 - 16]
-	ldp	q4, q5, [\state, #-16 * 4 - 16]
-	ldp	q2, q3, [\state, #-16 * 2 - 16]
-	ldp	q0, q1, [\state, #-16 * 0 - 16]
-0:
-.endm
diff --git a/arch/arm64/include/asm/neon.h b/arch/arm64/include/asm/neon.h
index 5368bd04fe7b..fb9d137256a6 100644
--- a/arch/arm64/include/asm/neon.h
+++ b/arch/arm64/include/asm/neon.h
@@ -16,9 +16,7 @@
 
 #define cpu_has_neon()		system_supports_fpsimd()
 
-#define kernel_neon_begin()	kernel_neon_begin_partial(32)
-
-void kernel_neon_begin_partial(u32 num_regs);
+void kernel_neon_begin(void);
 void kernel_neon_end(void);
 
 #endif /* ! __ASM_NEON_H */
diff --git a/arch/arm64/include/asm/simd.h b/arch/arm64/include/asm/simd.h
index 96959b52afae..5a1a927b74a2 100644
--- a/arch/arm64/include/asm/simd.h
+++ b/arch/arm64/include/asm/simd.h
@@ -9,15 +9,46 @@
 #ifndef __ASM_SIMD_H
 #define __ASM_SIMD_H
 
+#include <linux/compiler.h>
+#include <linux/percpu.h>
+#include <linux/preempt.h>
 #include <linux/types.h>
 
+#ifdef CONFIG_KERNEL_MODE_NEON
+
+DECLARE_PER_CPU(bool, kernel_neon_busy);
+
 /*
  * may_use_simd - whether it is allowable at this time to issue SIMD
  *                instructions or access the SIMD register file
+ *
+ * Callers must not assume that the result remains true beyond the next
+ * preempt_enable() or return from softirq context.
  */
 static __must_check inline bool may_use_simd(void)
 {
-	return true;
+	/*
+	 * The raw_cpu_read() is racy if called with preemption enabled.
+	 * This is not a bug: kernel_neon_busy is only set when
+	 * preemption is disabled, so we cannot migrate to another CPU
+	 * while it is set, nor can we migrate to a CPU where it is set.
+	 * So, if we find it clear on some CPU then we're guaranteed to
+	 * find it clear on any CPU we could migrate to.
+	 *
+	 * If we are in between kernel_neon_begin()...kernel_neon_end(),
+	 * the flag will be set, but preemption is also disabled, so we
+	 * can't migrate to another CPU and spuriously see it become
+	 * false.
+	 */
+	return !in_irq() && !in_nmi() && !raw_cpu_read(kernel_neon_busy);
 }
 
+#else /* ! CONFIG_KERNEL_MODE_NEON */
+
+static __must_check inline bool may_use_simd(void) {
+	return false;
+}
+
+#endif /* ! CONFIG_KERNEL_MODE_NEON */
+
 #endif