3 files changed, 120 insertions, 115 deletions

diff --git a/arch/x86/kernel/fpu/core.c b/arch/x86/kernel/fpu/core.c
index d25097c3fc1d..0b1b9abd4d5f 100644
--- a/arch/x86/kernel/fpu/core.c
+++ b/arch/x86/kernel/fpu/core.c
@@ -114,6 +114,10 @@ void __kernel_fpu_begin(void)
 	kernel_fpu_disable();
 
 	if (fpu->fpregs_active) {
+		/*
+		 * Ignore return value -- we don't care if reg state
+		 * is clobbered.
+		 */
 		copy_fpregs_to_fpstate(fpu);
 	} else {
 		this_cpu_write(fpu_fpregs_owner_ctx, NULL);
@@ -189,8 +193,12 @@ void fpu__save(struct fpu *fpu)
 
 	preempt_disable();
 	if (fpu->fpregs_active) {
-		if (!copy_fpregs_to_fpstate(fpu))
-			fpregs_deactivate(fpu);
+		if (!copy_fpregs_to_fpstate(fpu)) {
+			if (use_eager_fpu())
+				copy_kernel_to_fpregs(&fpu->state);
+			else
+				fpregs_deactivate(fpu);
+		}
 	}
 	preempt_enable();
 }
@@ -223,14 +231,15 @@ void fpstate_init(union fpregs_state *state)
 }
 EXPORT_SYMBOL_GPL(fpstate_init);
 
-/*
- * Copy the current task's FPU state to a new task's FPU context.
- *
- * In both the 'eager' and the 'lazy' case we save hardware registers
- * directly to the destination buffer.
- */
-static void fpu_copy(struct fpu *dst_fpu, struct fpu *src_fpu)
+int fpu__copy(struct fpu *dst_fpu, struct fpu *src_fpu)
 {
+	dst_fpu->counter = 0;
+	dst_fpu->fpregs_active = 0;
+	dst_fpu->last_cpu = -1;
+
+	if (!src_fpu->fpstate_active || !cpu_has_fpu)
+		return 0;
+
 	WARN_ON_FPU(src_fpu != &current->thread.fpu);
 
 	/*
@@ -243,10 +252,9 @@ static void fpu_copy(struct fpu *dst_fpu, struct fpu *src_fpu)
 	/*
 	 * Save current FPU registers directly into the child
 	 * FPU context, without any memory-to-memory copying.
-	 *
-	 * If the FPU context got destroyed in the process (FNSAVE
-	 * done on old CPUs) then copy it back into the source
-	 * context and mark the current task for lazy restore.
+	 * In lazy mode, if the FPU context isn't loaded into
+	 * fpregs, CR0.TS will be set and do_device_not_available
+	 * will load the FPU context.
 	 *
 	 * We have to do all this with preemption disabled,
 	 * mostly because of the FNSAVE case, because in that
@@ -259,19 +267,13 @@ static void fpu_copy(struct fpu *dst_fpu, struct fpu *src_fpu)
 	preempt_disable();
 	if (!copy_fpregs_to_fpstate(dst_fpu)) {
 		memcpy(&src_fpu->state, &dst_fpu->state, xstate_size);
-		fpregs_deactivate(src_fpu);
+
+		if (use_eager_fpu())
+			copy_kernel_to_fpregs(&src_fpu->state);
+		else
+			fpregs_deactivate(src_fpu);
 	}
 	preempt_enable();
-}
-
-int fpu__copy(struct fpu *dst_fpu, struct fpu *src_fpu)
-{
-	dst_fpu->counter = 0;
-	dst_fpu->fpregs_active = 0;
-	dst_fpu->last_cpu = -1;
-
-	if (src_fpu->fpstate_active && cpu_has_fpu)
-		fpu_copy(dst_fpu, src_fpu);
 
 	return 0;
 }
@@ -409,8 +411,10 @@ static inline void copy_init_fpstate_to_fpregs(void)
 {
 	if (use_xsave())
 		copy_kernel_to_xregs(&init_fpstate.xsave, -1);
-	else
+	else if (static_cpu_has(X86_FEATURE_FXSR))
 		copy_kernel_to_fxregs(&init_fpstate.fxsave);
+	else
+		copy_kernel_to_fregs(&init_fpstate.fsave);
 }
 
 /*
@@ -423,7 +427,7 @@ void fpu__clear(struct fpu *fpu)
 {
 	WARN_ON_FPU(fpu != &current->thread.fpu); /* Almost certainly an anomaly */
 
-	if (!use_eager_fpu()) {
+	if (!use_eager_fpu() || !static_cpu_has(X86_FEATURE_FPU)) {
 		/* FPU state will be reallocated lazily at the first use. */
 		fpu__drop(fpu);
 	} else {
diff --git a/arch/x86/kernel/fpu/init.c b/arch/x86/kernel/fpu/init.c
index 7b2978ab30df..54c86fffbf9f 100644
--- a/arch/x86/kernel/fpu/init.c
+++ b/arch/x86/kernel/fpu/init.c
@@ -3,8 +3,11 @@
  */
 #include <asm/fpu/internal.h>
 #include <asm/tlbflush.h>
+#include <asm/setup.h>
+#include <asm/cmdline.h>
 
 #include <linux/sched.h>
+#include <linux/init.h>
 
 /*
  * Initialize the TS bit in CR0 according to the style of context-switches
@@ -75,13 +78,15 @@ static void fpu__init_system_early_generic(struct cpuinfo_x86 *c)
 	cr0 &= ~(X86_CR0_TS | X86_CR0_EM);
 	write_cr0(cr0);
 
-	asm volatile("fninit ; fnstsw %0 ; fnstcw %1"
-		     : "+m" (fsw), "+m" (fcw));
+	if (!test_bit(X86_FEATURE_FPU, (unsigned long *)cpu_caps_cleared)) {
+		asm volatile("fninit ; fnstsw %0 ; fnstcw %1"
+			     : "+m" (fsw), "+m" (fcw));
 
-	if (fsw == 0 && (fcw & 0x103f) == 0x003f)
-		set_cpu_cap(c, X86_FEATURE_FPU);
-	else
-		clear_cpu_cap(c, X86_FEATURE_FPU);
+		if (fsw == 0 && (fcw & 0x103f) == 0x003f)
+			set_cpu_cap(c, X86_FEATURE_FPU);
+		else
+			clear_cpu_cap(c, X86_FEATURE_FPU);
+	}
 
 #ifndef CONFIG_MATH_EMULATION
 	if (!cpu_has_fpu) {
@@ -129,7 +134,7 @@ static void __init fpu__init_system_generic(void)
 	 * Set up the legacy init FPU context. (xstate init might overwrite this
 	 * with a more modern format, if the CPU supports it.)
 	 */
-	fpstate_init_fxstate(&init_fpstate.fxsave);
+	fpstate_init(&init_fpstate);
 
 	fpu__init_system_mxcsr();
 }
@@ -257,7 +262,10 @@ static void __init fpu__init_system_xstate_size_legacy(void)
  * not only saved the restores along the way, but we also have the
  * FPU ready to be used for the original task.
  *
- * 'eager' switching is used on modern CPUs, there we switch the FPU
+ * 'lazy' is deprecated because it's almost never a performance win
+ * and it's much more complicated than 'eager'.
+ *
+ * 'eager' switching is by default on all CPUs, there we switch the FPU
  * state during every context switch, regardless of whether the task
  * has used FPU instructions in that time slice or not. This is done
  * because modern FPU context saving instructions are able to optimize
@@ -268,22 +276,48 @@ static void __init fpu__init_system_xstate_size_legacy(void)
  *   to use 'eager' restores, if we detect that a task is using the FPU
  *   frequently. See the fpu->counter logic in fpu/internal.h for that. ]
  */
-static enum { AUTO, ENABLE, DISABLE } eagerfpu = AUTO;
+static enum { ENABLE, DISABLE } eagerfpu = ENABLE;
 
-static int __init eager_fpu_setup(char *s)
+/*
+ * Find supported xfeatures based on cpu features and command-line input.
+ * This must be called after fpu__init_parse_early_param() is called and
+ * xfeatures_mask is enumerated.
+ */
+u64 __init fpu__get_supported_xfeatures_mask(void)
 {
-	if (!strcmp(s, "on"))
-		eagerfpu = ENABLE;
-	else if (!strcmp(s, "off"))
-		eagerfpu = DISABLE;
-	else if (!strcmp(s, "auto"))
-		eagerfpu = AUTO;
-	return 1;
+	/* Support all xfeatures known to us */
+	if (eagerfpu != DISABLE)
+		return XCNTXT_MASK;
+
+	/* Warning of xfeatures being disabled for no eagerfpu mode */
+	if (xfeatures_mask & XFEATURE_MASK_EAGER) {
+		pr_err("x86/fpu: eagerfpu switching disabled, disabling the following xstate features: 0x%llx.\n",
+			xfeatures_mask & XFEATURE_MASK_EAGER);
+	}
+
+	/* Return a mask that masks out all features requiring eagerfpu mode */
+	return ~XFEATURE_MASK_EAGER;
+}
+
+/*
+ * Disable features dependent on eagerfpu.
+ */
+static void __init fpu__clear_eager_fpu_features(void)
+{
+	setup_clear_cpu_cap(X86_FEATURE_MPX);
 }
-__setup("eagerfpu=", eager_fpu_setup);
 
 /*
  * Pick the FPU context switching strategy:
+ *
+ * When eagerfpu is AUTO or ENABLE, we ensure it is ENABLE if either of
+ * the following is true:
+ *
+ * (1) the cpu has xsaveopt, as it has the optimization and doing eager
+ *     FPU switching has a relatively low cost compared to a plain xsave;
+ * (2) the cpu has xsave features (e.g. MPX) that depend on eager FPU
+ *     switching. Should the kernel boot with noxsaveopt, we support MPX
+ *     with eager FPU switching at a higher cost.
  */
 static void __init fpu__init_system_ctx_switch(void)
 {
@@ -295,19 +329,11 @@ static void __init fpu__init_system_ctx_switch(void)
 	WARN_ON_FPU(current->thread.fpu.fpstate_active);
 	current_thread_info()->status = 0;
 
-	/* Auto enable eagerfpu for xsaveopt */
 	if (boot_cpu_has(X86_FEATURE_XSAVEOPT) && eagerfpu != DISABLE)
 		eagerfpu = ENABLE;
 
-	if (xfeatures_mask & XFEATURE_MASK_EAGER) {
-		if (eagerfpu == DISABLE) {
-			pr_err("x86/fpu: eagerfpu switching disabled, disabling the following xstate features: 0x%llx.\n",
-			       xfeatures_mask & XFEATURE_MASK_EAGER);
-			xfeatures_mask &= ~XFEATURE_MASK_EAGER;
-		} else {
-			eagerfpu = ENABLE;
-		}
-	}
+	if (xfeatures_mask & XFEATURE_MASK_EAGER)
+		eagerfpu = ENABLE;
 
 	if (eagerfpu == ENABLE)
 		setup_force_cpu_cap(X86_FEATURE_EAGER_FPU);
@@ -316,11 +342,42 @@ static void __init fpu__init_system_ctx_switch(void)
 }
 
 /*
+ * We parse fpu parameters early because fpu__init_system() is executed
+ * before parse_early_param().
+ */
+static void __init fpu__init_parse_early_param(void)
+{
+	if (cmdline_find_option_bool(boot_command_line, "eagerfpu=off")) {
+		eagerfpu = DISABLE;
+		fpu__clear_eager_fpu_features();
+	}
+
+	if (cmdline_find_option_bool(boot_command_line, "no387"))
+		setup_clear_cpu_cap(X86_FEATURE_FPU);
+
+	if (cmdline_find_option_bool(boot_command_line, "nofxsr")) {
+		setup_clear_cpu_cap(X86_FEATURE_FXSR);
+		setup_clear_cpu_cap(X86_FEATURE_FXSR_OPT);
+		setup_clear_cpu_cap(X86_FEATURE_XMM);
+	}
+
+	if (cmdline_find_option_bool(boot_command_line, "noxsave"))
+		fpu__xstate_clear_all_cpu_caps();
+
+	if (cmdline_find_option_bool(boot_command_line, "noxsaveopt"))
+		setup_clear_cpu_cap(X86_FEATURE_XSAVEOPT);
+
+	if (cmdline_find_option_bool(boot_command_line, "noxsaves"))
+		setup_clear_cpu_cap(X86_FEATURE_XSAVES);
+}
+
+/*
  * Called on the boot CPU once per system bootup, to set up the initial
  * FPU state that is later cloned into all processes:
  */
 void __init fpu__init_system(struct cpuinfo_x86 *c)
 {
+	fpu__init_parse_early_param();
 	fpu__init_system_early_generic(c);
 
 	/*
@@ -344,62 +401,3 @@ void __init fpu__init_system(struct cpuinfo_x86 *c)
 
 	fpu__init_system_ctx_switch();
 }
-
-/*
- * Boot parameter to turn off FPU support and fall back to math-emu:
- */
-static int __init no_387(char *s)
-{
-	setup_clear_cpu_cap(X86_FEATURE_FPU);
-	return 1;
-}
-__setup("no387", no_387);
-
-/*
- * Disable all xstate CPU features:
- */
-static int __init x86_noxsave_setup(char *s)
-{
-	if (strlen(s))
-		return 0;
-
-	fpu__xstate_clear_all_cpu_caps();
-
-	return 1;
-}
-__setup("noxsave", x86_noxsave_setup);
-
-/*
- * Disable the XSAVEOPT instruction specifically:
- */
-static int __init x86_noxsaveopt_setup(char *s)
-{
-	setup_clear_cpu_cap(X86_FEATURE_XSAVEOPT);
-
-	return 1;
-}
-__setup("noxsaveopt", x86_noxsaveopt_setup);
-
-/*
- * Disable the XSAVES instruction:
- */
-static int __init x86_noxsaves_setup(char *s)
-{
-	setup_clear_cpu_cap(X86_FEATURE_XSAVES);
-
-	return 1;
-}
-__setup("noxsaves", x86_noxsaves_setup);
-
-/*
- * Disable FX save/restore and SSE support:
- */
-static int __init x86_nofxsr_setup(char *s)
-{
-	setup_clear_cpu_cap(X86_FEATURE_FXSR);
-	setup_clear_cpu_cap(X86_FEATURE_FXSR_OPT);
-	setup_clear_cpu_cap(X86_FEATURE_XMM);
-
-	return 1;
-}
-__setup("nofxsr", x86_nofxsr_setup);
diff --git a/arch/x86/kernel/fpu/xstate.c b/arch/x86/kernel/fpu/xstate.c
index 40f100285984..6e8354f5a593 100644
--- a/arch/x86/kernel/fpu/xstate.c
+++ b/arch/x86/kernel/fpu/xstate.c
@@ -51,7 +51,11 @@ void fpu__xstate_clear_all_cpu_caps(void)
 	setup_clear_cpu_cap(X86_FEATURE_AVX512PF);
 	setup_clear_cpu_cap(X86_FEATURE_AVX512ER);
 	setup_clear_cpu_cap(X86_FEATURE_AVX512CD);
+	setup_clear_cpu_cap(X86_FEATURE_AVX512DQ);
+	setup_clear_cpu_cap(X86_FEATURE_AVX512BW);
+	setup_clear_cpu_cap(X86_FEATURE_AVX512VL);
 	setup_clear_cpu_cap(X86_FEATURE_MPX);
+	setup_clear_cpu_cap(X86_FEATURE_XGETBV1);
 }
 
 /*
@@ -632,8 +636,7 @@ void __init fpu__init_system_xstate(void)
 		BUG();
 	}
 
-	/* Support only the state known to the OS: */
-	xfeatures_mask = xfeatures_mask & XCNTXT_MASK;
+	xfeatures_mask &= fpu__get_supported_xfeatures_mask();
 
 	/* Enable xstate instructions to be able to continue with initialization: */
 	fpu__init_cpu_xstate();