Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1536577

[RFC PATCH 5/6] x86/fpu: Fix CPUID-less FPU detection

Show all headers | View raw

The old code didn't work at all because it adjusted the current caps
instead of the forced caps.  Anything it did would be undone later
during cpu identification.  Fix that and, while we're at it, improve
the logging and don't bother running it if CPUID is available.

Signed-off-by: Andy Lutomirski <luto@kernel.org>
---
 arch/x86/kernel/fpu/init.c | 28 ++++++++++++++++------------
 1 file changed, 16 insertions(+), 12 deletions(-)

diff --git a/arch/x86/kernel/fpu/init.c b/arch/x86/kernel/fpu/init.c
index 60dece392b3a..75e1bf3b0319 100644
--- a/arch/x86/kernel/fpu/init.c
+++ b/arch/x86/kernel/fpu/init.c
@@ -50,29 +50,33 @@ void fpu__init_cpu(void)
 
 /*
  * The earliest FPU detection code.
- *
- * Set the X86_FEATURE_FPU CPU-capability bit based on
- * trying to execute an actual sequence of FPU instructions:
  */
 static void fpu__init_system_early_generic(struct cpuinfo_x86 *c)
 {
-	unsigned long cr0;
-	u16 fsw, fcw;
+	if (!boot_cpu_has(X86_FEATURE_CPUID) &&
+	    !test_bit(X86_FEATURE_FPU, (unsigned long *)cpu_caps_cleared)) {
+		/*
+		 * Set the X86_FEATURE_FPU CPU-capability bit based on
+		 * trying to execute an actual sequence of FPU instructions:
+		 */
 
-	fsw = fcw = 0xffff;
+		unsigned long cr0;
+		u16 fsw, fcw;
 
-	cr0 = read_cr0();
-	cr0 &= ~(X86_CR0_TS | X86_CR0_EM);
-	write_cr0(cr0);
+		fsw = fcw = 0xffff;
+
+		cr0 = read_cr0();
+		cr0 &= ~(X86_CR0_TS | X86_CR0_EM);
+		write_cr0(cr0);
 
-	if (!test_bit(X86_FEATURE_FPU, (unsigned long *)cpu_caps_cleared)) {
 		asm volatile("fninit ; fnstsw %0 ; fnstcw %1"
 			     : "+m" (fsw), "+m" (fcw));
+		pr_info("x86/fpu: FSW=0x%04hx FCW=0x%04hx\n", fsw, fcw);
 
 		if (fsw == 0 && (fcw & 0x103f) == 0x003f)
-			set_cpu_cap(c, X86_FEATURE_FPU);
+			setup_force_cpu_cap(X86_FEATURE_FPU);
 		else
-			clear_cpu_cap(c, X86_FEATURE_FPU);
+			setup_clear_cpu_cap(X86_FEATURE_FPU);
 	}
 
 #ifndef CONFIG_MATH_EMULATION
-- 
2.9.3

Back to linux.kernel | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread

Thread

[RFC PATCH 0/6] CPUID and FPU fixes Andy Lutomirski <luto@kernel.org> - 2016-12-06 02:10 +0100
  [RFC PATCH 4/6] x86/fpu: Fix "x86/fpu: Legacy x87 FPU detected" message Andy Lutomirski <luto@kernel.org> - 2016-12-06 02:10 +0100
    Re: [RFC PATCH 4/6] x86/fpu: Fix "x86/fpu: Legacy x87 FPU detected"  message Borislav Petkov <bp@alien8.de> - 2016-12-06 10:30 +0100
      Re: [RFC PATCH 4/6] x86/fpu: Fix "x86/fpu: Legacy x87 FPU detected" message Andy Lutomirski <luto@amacapital.net> - 2016-12-06 19:10 +0100
        Re: [RFC PATCH 4/6] x86/fpu: Fix "x86/fpu: Legacy x87 FPU detected"  message Borislav Petkov <bp@alien8.de> - 2016-12-06 19:10 +0100
  [RFC PATCH 5/6] x86/fpu: Fix CPUID-less FPU detection Andy Lutomirski <luto@kernel.org> - 2016-12-06 02:10 +0100
    Re: [RFC PATCH 5/6] x86/fpu: Fix CPUID-less FPU detection Borislav Petkov <bp@alien8.de> - 2016-12-06 10:50 +0100
      Re: [RFC PATCH 5/6] x86/fpu: Fix CPUID-less FPU detection Andy Lutomirski <luto@amacapital.net> - 2016-12-06 19:00 +0100
        Re: [RFC PATCH 5/6] x86/fpu: Fix CPUID-less FPU detection Borislav Petkov <bp@alien8.de> - 2016-12-07 10:50 +0100
  [RFC PATCH 6/6] x86/fpu: Fix the "Giving up, no FPU found" test Andy Lutomirski <luto@kernel.org> - 2016-12-06 02:10 +0100
  [RFC PATCH 3/6] x86/cpu: Re-apply forced caps every time cpu caps are re-read Andy Lutomirski <luto@kernel.org> - 2016-12-06 02:10 +0100
    Re: [RFC PATCH 3/6] x86/cpu: Re-apply forced caps every time cpu  caps are re-read Borislav Petkov <bp@alien8.de> - 2016-12-06 10:00 +0100
  [RFC PATCH 2/6] x86/cpu: Factor out application of forced cpu caps Andy Lutomirski <luto@kernel.org> - 2016-12-06 02:10 +0100
    Re: [RFC PATCH 2/6] x86/cpu: Factor out application of forced cpu  caps Borislav Petkov <bp@alien8.de> - 2016-12-06 09:40 +0100

csiph-web