[PATCH] x86/ucode/intel: Improve description for gathering the microcode revision

Andrew Cooper posted 1 patch 2 weeks ago
Patches applied successfully (tree, apply log)
git fetch https://github.com/patchew-project/xen tags/patchew/20201012142523.17652-1-andrew.cooper3@citrix.com
xen/arch/x86/cpu/microcode/intel.c | 26 +++++++++++++++-----------
1 file changed, 15 insertions(+), 11 deletions(-)

[PATCH] x86/ucode/intel: Improve description for gathering the microcode revision

Posted by Andrew Cooper 2 weeks ago
Obtaining the microcode revision on Intel CPUs is complicated for backwards
compatibility reasons.  Update apply_microcode() to use a slightly more
efficient CPUID invocation, now that the documentation has been updated to
confirm that any CPUID instruction is fine, not just CPUID.1

Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
---
CC: Jan Beulich <JBeulich@suse.com>
CC: Roger Pau Monné <roger.pau@citrix.com>
CC: Wei Liu <wl@xen.org>
---
 xen/arch/x86/cpu/microcode/intel.c | 26 +++++++++++++++-----------
 1 file changed, 15 insertions(+), 11 deletions(-)

diff --git a/xen/arch/x86/cpu/microcode/intel.c b/xen/arch/x86/cpu/microcode/intel.c
index d9bb1bc10e..72c07fcd1d 100644
--- a/xen/arch/x86/cpu/microcode/intel.c
+++ b/xen/arch/x86/cpu/microcode/intel.c
@@ -126,13 +126,16 @@ static void collect_cpu_info(void)
     rdmsrl(MSR_IA32_PLATFORM_ID, msr_content);
     csig->pf = 1 << ((msr_content >> 50) & 7);
 
-    wrmsrl(MSR_IA32_UCODE_REV, 0x0ULL);
-    /* As documented in the SDM: Do a CPUID 1 here */
+    /*
+     * Obtaining the microcode version involves writing 0 to the "read only"
+     * UCODE_REV MSR, executing any CPUID instruction, after which a nonzero
+     * revision should appear.
+     */
+    wrmsrl(MSR_IA32_UCODE_REV, 0);
     csig->sig = cpuid_eax(1);
-
-    /* get the current revision from MSR 0x8B */
     rdmsrl(MSR_IA32_UCODE_REV, msr_content);
-    csig->rev = (uint32_t)(msr_content >> 32);
+    csig->rev = msr_content >> 32;
+
     pr_debug("microcode: collect_cpu_info : sig=%#x, pf=%#x, rev=%#x\n",
              csig->sig, csig->pf, csig->rev);
 }
@@ -270,14 +273,15 @@ static int apply_microcode(const struct microcode_patch *patch)
 
     wbinvd();
 
-    /* write microcode via MSR 0x79 */
     wrmsrl(MSR_IA32_UCODE_WRITE, (unsigned long)patch->data);
-    wrmsrl(MSR_IA32_UCODE_REV, 0x0ULL);
 
-    /* As documented in the SDM: Do a CPUID 1 here */
-    cpuid_eax(1);
-
-    /* get the current revision from MSR 0x8B */
+    /*
+     * Obtaining the microcode version involves writing 0 to the "read only"
+     * UCODE_REV MSR, executing any CPUID instruction, after which a nonzero
+     * revision should appear.
+     */
+    wrmsrl(MSR_IA32_UCODE_REV, 0);
+    cpuid_eax(0);
     rdmsrl(MSR_IA32_UCODE_REV, msr_content);
     sig->rev = rev = msr_content >> 32;
 
-- 
2.11.0


Re: [PATCH] x86/ucode/intel: Improve description for gathering the microcode revision

Posted by Jan Beulich 2 weeks ago
On 12.10.2020 16:25, Andrew Cooper wrote:
> Obtaining the microcode revision on Intel CPUs is complicated for backwards
> compatibility reasons.  Update apply_microcode() to use a slightly more
> efficient CPUID invocation, now that the documentation has been updated to
> confirm that any CPUID instruction is fine, not just CPUID.1
> 
> Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>

Reviewed-by: Jan Beulich <jbeulich@suse.com>