Skip to contentSkip to conversation

MSR fixes and cleanups after last round of MSR cleanups

PATCH Applied Xin Li (Intel) 4 participants 22 messages May 12, 2025 – May 21, 2025

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

cover MSR fixes and cleanups after last round of MSR cleanups

These patches:

1) remove a superfluous inclusion of <asm/asm.h> accidently added to drivers/acpi/processor_throttling.c in commit:

     efef7f184f2e ("x86/msr: Add explicit includes of <asm/msr.h>").

2) Fix uninitialized symbol 'err' introduced by:

     d815da84fdd0 ("x86/msr: Change the function type of native_read_msr_safe()").

3) Convert a native_wrmsr() use to native_wrmsrq() in arch/x86/coco/sev/core.c.

Xin Li (Intel) (3):
x86/msr: Remove a superfluous inclusion of <asm/asm.h> x86/xen/msr: Fix uninitialized symbol 'err' x86/msr: Convert a native_wrmsr() use to native_wrmsrq()

 arch/x86/coco/sev/core.c            | 7 +------
 arch/x86/xen/enlighten_pv.c         | 5 ++++-
 drivers/acpi/processor_throttling.c | 1 -
 3 files changed, 5 insertions(+), 8 deletions(-)

base-commit: 9cf78722003178b09c409df9aafe9d79e5b9a74e

Signature · 1 line
2.49.0
#1 lore raw

PATCH 1/3 x86/msr: Remove a superfluous inclusion of <asm/asm.h>

The following commit:

efef7f184f2e ("x86/msr: Add explicit includes of <asm/msr.h>")

added a superfluous inclusion of <asm/asm.h> to

drivers/acpi/processor_throttling.c.

Remove it.

---
 drivers/acpi/processor_throttling.c | 1 -
 1 file changed, 1 deletion(-)
drivers/acpi/processor_throttling.c +0 -1
index ecd7fe256153..d1541a386fbc 100644
@@ -21,7 +21,6 @@
 #include <linux/uaccess.h>
 #include <acpi/processor.h>
 #include <asm/io.h>
-#include <asm/asm.h>
 #ifdef CONFIG_X86
 #include <asm/msr.h>
 #endif
Signature · 1 line
2.49.0
merge bot Applied
#2 lore raw

[tip: x86/core] x86/msr: Remove a superfluous inclusion of <asm/asm.h>

The following commit has been merged into the x86/core branch of tip:

The following commit has been merged into the x86/core branch of tip:

Commit-ID:     9220aa8a6779b586ef11bcd5473d103f7cf60756
Gitweb:        https://git.kernel.org/tip/9220aa8a6779b586ef11bcd5473d103f7cf60756
Author:        Xin Li (Intel) <[email protected]>
AuthorDate:    Mon, 12 May 2025 01:45:50 -07:00
Committer:     Ingo Molnar <[email protected]>
CommitterDate: Sun, 18 May 2025 08:39:09 +02:00

x86/msr: Remove a superfluous inclusion of <asm/asm.h>

The following commit:

efef7f184f2e ("x86/msr: Add explicit includes of <asm/msr.h>")

added a superfluous inclusion of <asm/asm.h> to

drivers/acpi/processor_throttling.c.

Remove it.

Fixes efef7f184f2e ("x86/msr: Add explicit includes of <asm/msr.h>")
Cc H. Peter Anvin <[email protected]>
---
 drivers/acpi/processor_throttling.c | 1 -
 1 file changed, 1 deletion(-)
drivers/acpi/processor_throttling.c +0 -1
index ecd7fe2..d1541a3 100644
@@ -21,7 +21,6 @@
 #include <linux/uaccess.h>
 #include <acpi/processor.h>
 #include <asm/io.h>
-#include <asm/asm.h>
 #ifdef CONFIG_X86
 #include <asm/msr.h>
 #endif
#3 lore raw

PATCH 2/3 x86/xen/msr: Fix uninitialized symbol 'err'

In reply to Xin Li (Intel)

xen_read_msr_safe() currently passes an uninitialized argument err to xen_do_read_msr(). But as xen_do_read_msr() may not set the argument, xen_read_msr_safe() could return err with an unpredictable value.

To ensure correctness, initialize err to 0 (representing success) in xen_read_msr_safe().

Because xen_read_msr_safe() is essentially a wrapper of xen_do_read_msr(), the latter should be responsible for initializing the value of *err to 0. Thus initialize *err to 0 in xen_do_read_msr().

Fixes 502ad6e5a619 ("x86/msr: Change the function type of native_read_msr_safe()")
---
 arch/x86/xen/enlighten_pv.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
arch/x86/xen/enlighten_pv.c +4 -1
index 3be38350f044..01f1d441347e 100644
@@ -1091,6 +1091,9 @@ static u64 xen_do_read_msr(u32 msr, int *err)
 {
 	u64 val = 0;	/* Avoid uninitialized value for safe variant. */
 
+	if (err)
+		*err = 0;
+
 	if (pmu_msr_chk_emulated(msr, &val, true))
 		return val;
 
@@ -1162,7 +1165,7 @@ static void xen_do_write_msr(u32 msr, u64 val, int *err)
 
 static int xen_read_msr_safe(u32 msr, u64 *val)
 {
-	int err;
+	int err = 0;
 
 	*val = xen_do_read_msr(msr, &err);
 	return err;
Signature · 1 line
2.49.0
#4 lore raw

Re: [PATCH v1 2/3] x86/xen/msr: Fix uninitialized symbol 'err'

Quoted 34 lines · Xin Li (Intel)

xen_read_msr_safe() currently passes an uninitialized argument err to xen_do_read_msr(). But as xen_do_read_msr() may not set the argument, xen_read_msr_safe() could return err with an unpredictable value.

To ensure correctness, initialize err to 0 (representing success) in xen_read_msr_safe().

Because xen_read_msr_safe() is essentially a wrapper of xen_do_read_msr(), the latter should be responsible for initializing the value of *err to 0. Thus initialize *err to 0 in xen_do_read_msr().

Fixes 502ad6e5a619 ("x86/msr: Change the function type of native_read_msr_safe()")
---
 arch/x86/xen/enlighten_pv.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
arch/x86/xen/enlighten_pv.c +3 -0
index 3be38350f044..01f1d441347e 100644
@@ -1091,6 +1091,9 @@ static u64 xen_do_read_msr(u32 msr, int *err)
 {
 	u64 val = 0;	/* Avoid uninitialized value for safe variant. */
 
+	if (err)
+		*err = 0;
+
 	if (pmu_msr_chk_emulated(msr, &val, true))
 		return val;
 

arch/x86/xen/enlighten_pv.c static void xen_do_write_msr(u32 msr, u64 val, int *err) · line 1171

Quoted 8 lines
arch/x86/xen/enlighten_pv.c +1 -1
@@ -1162,7 +1165,7 @@ static void xen_do_write_msr(u32 msr, u64 val, int *err)
 
 static int xen_read_msr_safe(u32 msr, u64 *val)
 {
-	int err;
+	int err = 0;
 
 	*val = xen_do_read_msr(msr, &err);
 	return err;

So why not initialize 'err' with 0 in both callers, xen_read_msr_safe() and xen_read_msr(), and avoid all the initialization trouble in xen_do_read_msr()?

Thanks,

Ingo

author
#5 lore raw

Re: [PATCH v1 2/3] x86/xen/msr: Fix uninitialized symbol 'err'

Quoted 43 lines · Ingo Molnar
Quoted 40 lines · Xin Li (Intel)

xen_read_msr_safe() currently passes an uninitialized argument err to xen_do_read_msr(). But as xen_do_read_msr() may not set the argument, xen_read_msr_safe() could return err with an unpredictable value.

To ensure correctness, initialize err to 0 (representing success) in xen_read_msr_safe().

Because xen_read_msr_safe() is essentially a wrapper of xen_do_read_msr(), the latter should be responsible for initializing the value of *err to 0. Thus initialize *err to 0 in xen_do_read_msr().

Fixes 502ad6e5a619 ("x86/msr: Change the function type of native_read_msr_safe()")
---
  arch/x86/xen/enlighten_pv.c | 5 ++++-
  1 file changed, 4 insertions(+), 1 deletion(-)
arch/x86/xen/enlighten_pv.c +4 -1
index 3be38350f044..01f1d441347e 100644
@@ -1091,6 +1091,9 @@ static u64 xen_do_read_msr(u32 msr, int *err)
  {
  	u64 val = 0;	/* Avoid uninitialized value for safe variant. */
 
+	if (err)
+		*err = 0;
+
  	if (pmu_msr_chk_emulated(msr, &val, true))
  		return val;
 
@@ -1162,7 +1165,7 @@ static void xen_do_write_msr(u32 msr, u64 val, int *err)
 
  static int xen_read_msr_safe(u32 msr, u64 *val)
  {
-	int err;
+	int err = 0;
 
Quoted 6 lines

arch/x86/xen/enlighten_pv.c static void xen_do_write_msr(u32 msr, u64 val, int *err) · line 1171

Quoted 2 lines
arch/x86/xen/enlighten_pv.c +0 -0
  	*val = xen_do_read_msr(msr, &err);
  	return err;

So why not initialize 'err' with 0 in both callers, xen_read_msr_safe() and xen_read_msr(), and avoid all the initialization trouble in xen_do_read_msr()?

Yeah, I should make the change in xen_read_msr() too.

However xen_do_read_msr() should be implemented in a defensive way to set *err properly as it's part of its return value. Actually it was so, but one of my previous cleanup patch removed it because err is no longer passed to pmu_msr_chk_emulated().

Thanks!
Xin

#6 lore raw

Re: [PATCH v1 2/3] x86/xen/msr: Fix uninitialized symbol 'err'

Quoted 47 lines · Xin Li
Quoted 37 lines · Ingo Molnar
Quoted 34 lines · Xin Li (Intel)

xen_read_msr_safe() currently passes an uninitialized argument err to xen_do_read_msr(). But as xen_do_read_msr() may not set the argument, xen_read_msr_safe() could return err with an unpredictable value.

To ensure correctness, initialize err to 0 (representing success) in xen_read_msr_safe().

Because xen_read_msr_safe() is essentially a wrapper of xen_do_read_msr(), the latter should be responsible for initializing the value of *err to 0. Thus initialize *err to 0 in xen_do_read_msr().

Fixes 502ad6e5a619 ("x86/msr: Change the function type of native_read_msr_safe()")
---
  arch/x86/xen/enlighten_pv.c | 5 ++++-
  1 file changed, 4 insertions(+), 1 deletion(-)
arch/x86/xen/enlighten_pv.c +3 -0
index 3be38350f044..01f1d441347e 100644
@@ -1091,6 +1091,9 @@ static u64 xen_do_read_msr(u32 msr, int *err)
  {
  	u64 val = 0;	/* Avoid uninitialized value for safe variant. */
+	if (err)
+		*err = 0;
+
  	if (pmu_msr_chk_emulated(msr, &val, true))
  		return val;
@@ -1162,7 +1165,7 @@ static void xen_do_write_msr(u32 msr, u64 val, int *err)
  static int xen_read_msr_safe(u32 msr, u64 *val)
  {
Quoted 8 lines

arch/x86/xen/enlighten_pv.c static void xen_do_write_msr(u32 msr, u64 val, int *err) · line 1171

Quoted 4 lines
arch/x86/xen/enlighten_pv.c +1 -1
-	int err;
+	int err = 0;
  	*val = xen_do_read_msr(msr, &err);
  	return err;

So why not initialize 'err' with 0 in both callers, xen_read_msr_safe() and xen_read_msr(), and avoid all the initialization trouble in xen_do_read_msr()?

Quoted 6 lines

Yeah, I should make the change in xen_read_msr() too.

However xen_do_read_msr() should be implemented in a defensive way to set *err properly as it's part of its return value. Actually it was so, but one of my previous cleanup patch removed it because err is no longer passed to pmu_msr_chk_emulated().

Maybe. It's up to Juergen though.

Thanks,

Ingo

maintainer
#7 lore raw

Re: [PATCH v1 2/3] x86/xen/msr: Fix uninitialized symbol 'err'

In reply to Xin Li
Quoted 47 lines · Xin Li
Quoted 37 lines · Ingo Molnar
Quoted 34 lines · Xin Li (Intel)

xen_read_msr_safe() currently passes an uninitialized argument err to xen_do_read_msr().  But as xen_do_read_msr() may not set the argument, xen_read_msr_safe() could return err with an unpredictable value.

To ensure correctness, initialize err to 0 (representing success) in xen_read_msr_safe().

Because xen_read_msr_safe() is essentially a wrapper of xen_do_read_msr(), the latter should be responsible for initializing the value of *err to 0. Thus initialize *err to 0 in xen_do_read_msr().

Fixes 502ad6e5a619 ("x86/msr: Change the function type of native_read_msr_safe()")

---
  arch/x86/xen/enlighten_pv.c | 5 ++++-

  1 file changed, 4 insertions(+), 1 deletion(-)
arch/x86/xen/enlighten_pv.c +0 -0
index 3be38350f044..01f1d441347e 100644
@@ -1091,6 +1091,9 @@ static u64 xen_do_read_msr(u32 msr, int *err)
   {

      u64 val = 0;    /* Avoid uninitialized value for safe variant. */ +    if (err)
+        *err = 0;
+
      if (pmu_msr_chk_emulated(msr, &val, true))
          return val;

arch/x86/xen/enlighten_pv.c +0 -0
@@ -1162,7 +1165,7 @@ static void xen_do_write_msr(u32 msr, u64 val, int *err)
   static int xen_read_msr_safe(u32 msr, u64 *val)
   {
Quoted 8 lines

arch/x86/xen/enlighten_pv.c static void xen_do_write_msr(u32 msr, u64 val, int *err) · line 1171

Quoted 4 lines
arch/x86/xen/enlighten_pv.c +1 -1
-    int err;
+    int err = 0;
       *val = xen_do_read_msr(msr, &err);
       return err;

So why not initialize 'err' with 0 in both callers, xen_read_msr_safe() and xen_read_msr(), and avoid all the initialization trouble in xen_do_read_msr()?

Quoted 6 lines

Yeah, I should make the change in xen_read_msr() too.

However xen_do_read_msr() should be implemented in a defensive way to set *err properly as it's part of its return value.  Actually it was so, but one of my previous cleanup patch removed it because err is no longer passed to pmu_msr_chk_emulated().

xen_do_read_msr() is usable only in enlighten_pv.c as it is static.

So I'd prefer to drop setting err to 0 in xen_do_read_msr() initially and to set err to 0 in all callers.

Juergen

author
#8 lore raw

Re: [PATCH v1 2/3] x86/xen/msr: Fix uninitialized symbol 'err'

Quoted 56 lines · Jürgen Groß
Quoted 48 lines · Xin Li
Quoted 38 lines · Ingo Molnar
Quoted 35 lines · Xin Li (Intel)

xen_read_msr_safe() currently passes an uninitialized argument err to xen_do_read_msr().  But as xen_do_read_msr() may not set the argument, xen_read_msr_safe() could return err with an unpredictable value.

To ensure correctness, initialize err to 0 (representing success) in xen_read_msr_safe().

Because xen_read_msr_safe() is essentially a wrapper of xen_do_read_msr(), the latter should be responsible for initializing the value of *err to 0. Thus initialize *err to 0 in xen_do_read_msr().

Fixes 502ad6e5a619 ("x86/msr: Change the function type of native_read_msr_safe()")

---
  arch/x86/xen/enlighten_pv.c | 5 ++++-

  1 file changed, 4 insertions(+), 1 deletion(-)
arch/x86/xen/enlighten_pv.c +0 -0
index 3be38350f044..01f1d441347e 100644
@@ -1091,6 +1091,9 @@ static u64 xen_do_read_msr(u32 msr, int *err)
   {

      u64 val = 0;    /* Avoid uninitialized value for safe variant. */ +    if (err)
+        *err = 0;
+
      if (pmu_msr_chk_emulated(msr, &val, true))
          return val;

arch/x86/xen/enlighten_pv.c +0 -0
@@ -1162,7 +1165,7 @@ static void xen_do_write_msr(u32 msr, u64 val, 
 int *err)
   static int xen_read_msr_safe(u32 msr, u64 *val)
   {
Quoted 8 lines

arch/x86/xen/enlighten_pv.c static void xen_do_write_msr(u32 msr, u64 val, int *err) · line 1171

Quoted 4 lines
arch/x86/xen/enlighten_pv.c +1 -1
-    int err;
+    int err = 0;
       *val = xen_do_read_msr(msr, &err);
       return err;

So why not initialize 'err' with 0 in both callers, xen_read_msr_safe() and xen_read_msr(), and avoid all the initialization trouble in xen_do_read_msr()?

Quoted 6 lines

Yeah, I should make the change in xen_read_msr() too.

However xen_do_read_msr() should be implemented in a defensive way to set *err properly as it's part of its return value.  Actually it was so, but one of my previous cleanup patch removed it because err is no longer passed to pmu_msr_chk_emulated().

Quoted 4 lines

xen_do_read_msr() is usable only in enlighten_pv.c as it is static.

So I'd prefer to drop setting err to 0 in xen_do_read_msr() initially and to set err to 0 in all callers.

Okay, I will send v1A to address this comment then.

Thanks!
Xin

Quoted 1 line

Juergen

#9 lore raw

PATCH 2/3 x86/xen/msr: Fix uninitialized symbol 'err'

In reply to Jürgen Groß

xen_read_msr_safe() currently passes an uninitialized argument err to xen_do_read_msr(). But as xen_do_read_msr() may not set the argument, xen_read_msr_safe() could return err with an unpredictable value.

To ensure correctness, initialize err to 0 (representing success) in xen_read_msr_safe().

Do the same in xen_read_msr(), even err is not used after being passed to xen_do_read_msr().

Fixes d815da84fdd0 ("x86/msr: Change the function type of native_read_msr_safe()"

---

Change in v1A:
*) Drop setting err to 0 in xen_do_read_msr() initially and set err to
0 in all callers (Jürgen Groß).

---
 arch/x86/xen/enlighten_pv.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
arch/x86/xen/enlighten_pv.c +2 -2
index 7f9ded1bc707..26bbaf4b7330 100644
@@ -1162,7 +1162,7 @@ static void xen_do_write_msr(u32 msr, u64 val, int *err)
 
 static int xen_read_msr_safe(u32 msr, u64 *val)
 {
-	int err;
+	int err = 0;
 
 	*val = xen_do_read_msr(msr, &err);
 	return err;
@@ -1179,7 +1179,7 @@ static int xen_write_msr_safe(u32 msr, u64 val)
 
 static u64 xen_read_msr(u32 msr)
 {
-	int err;
+	int err = 0;
 
 	return xen_do_read_msr(msr, xen_msr_safe ? &err : NULL);
 }
Signature · 1 line
2.49.0
maintainer Reviewed-by on v1 2/3
#10 lore raw

Re: [PATCH v1A 2/3] x86/xen/msr: Fix uninitialized symbol 'err'

Quoted 7 lines · Xin Li (Intel)

xen_read_msr_safe() currently passes an uninitialized argument err to xen_do_read_msr(). But as xen_do_read_msr() may not set the argument, xen_read_msr_safe() could return err with an unpredictable value.

To ensure correctness, initialize err to 0 (representing success) in xen_read_msr_safe().

Quoted 7 lines

Do the same in xen_read_msr(), even err is not used after being passed to xen_do_read_msr().

Fixes d815da84fdd0 ("x86/msr: Change the function type of native_read_msr_safe()"

Juergen

merge bot Applied
#11 lore raw

[tip: x86/core] x86/xen/msr: Fix uninitialized variable 'err'

In reply to Xin Li (Intel)
The following commit has been merged into the x86/core branch of tip:

The following commit has been merged into the x86/core branch of tip:

Commit-ID:     54c2c688cd9305bdbab4883b9da6ff63f4deca5d
Gitweb:        https://git.kernel.org/tip/54c2c688cd9305bdbab4883b9da6ff63f4deca5d
Author:        Xin Li (Intel) <[email protected]>
AuthorDate:    Sat, 17 May 2025 09:57:12 -07:00
Committer:     Ingo Molnar <[email protected]>
CommitterDate: Sun, 18 May 2025 08:39:16 +02:00

x86/xen/msr: Fix uninitialized variable 'err'

xen_read_msr_safe() currently passes an uninitialized argument 'err' to xen_do_read_msr(). But as xen_do_read_msr() may not set the argument, xen_read_msr_safe() could return err with an unpredictable value.

To ensure correctness, initialize err to 0 (representing success) in xen_read_msr_safe().

Do the same in xen_read_msr(), even err is not used after being passed to xen_do_read_msr().

---
 arch/x86/xen/enlighten_pv.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
arch/x86/xen/enlighten_pv.c +2 -2
index 7f9ded1..26bbaf4 100644
@@ -1162,7 +1162,7 @@ static void xen_do_write_msr(u32 msr, u64 val, int *err)
 
 static int xen_read_msr_safe(u32 msr, u64 *val)
 {
-	int err;
+	int err = 0;
 
 	*val = xen_do_read_msr(msr, &err);
 	return err;
@@ -1179,7 +1179,7 @@ static int xen_write_msr_safe(u32 msr, u64 val)
 
 static u64 xen_read_msr(u32 msr)
 {
-	int err;
+	int err = 0;
 
 	return xen_do_read_msr(msr, xen_msr_safe ? &err : NULL);
 }
merge bot Applied
#12 lore raw

[tip: x86/core] x86/xen/msr: Fix uninitialized variable 'err'

In reply to Xin Li (Intel)
The following commit has been merged into the x86/core branch of tip:

The following commit has been merged into the x86/core branch of tip:

Commit-ID:     e95534e107d2e9e136aa4d7cbededb3827e80074
Gitweb:        https://git.kernel.org/tip/e95534e107d2e9e136aa4d7cbededb3827e80074
Author:        Xin Li (Intel) <[email protected]>
AuthorDate:    Sat, 17 May 2025 09:57:12 -07:00
Committer:     Ingo Molnar <[email protected]>
CommitterDate: Wed, 21 May 2025 08:44:31 +02:00

x86/xen/msr: Fix uninitialized variable 'err'

xen_read_msr_safe() currently passes an uninitialized argument 'err' to xen_do_read_msr(). But as xen_do_read_msr() may not set the argument, xen_read_msr_safe() could return err with an unpredictable value.

To ensure correctness, initialize err to 0 (representing success) in xen_read_msr_safe().

Do the same in xen_read_msr(), even err is not used after being passed to xen_do_read_msr().

---
 arch/x86/xen/enlighten_pv.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
arch/x86/xen/enlighten_pv.c +2 -2
index 7f9ded1..26bbaf4 100644
@@ -1162,7 +1162,7 @@ static void xen_do_write_msr(u32 msr, u64 val, int *err)
 
 static int xen_read_msr_safe(u32 msr, u64 *val)
 {
-	int err;
+	int err = 0;
 
 	*val = xen_do_read_msr(msr, &err);
 	return err;
@@ -1179,7 +1179,7 @@ static int xen_write_msr_safe(u32 msr, u64 val)
 
 static u64 xen_read_msr(u32 msr)
 {
-	int err;
+	int err = 0;
 
 	return xen_do_read_msr(msr, xen_msr_safe ? &err : NULL);
 }
#13 lore raw

PATCH 3/3 x86/msr: Convert a native_wrmsr() use to native_wrmsrq()

In reply to Xin Li (Intel)

Convert a native_wrmsr() use to native_wrmsrq() to zap meaningless type conversions when a u64 MSR value is splitted into two u32.

---
 arch/x86/coco/sev/core.c | 7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)
arch/x86/coco/sev/core.c +1 -6
index ff82151f7718..b3ce6fc8b62d 100644
@@ -282,12 +282,7 @@ static inline u64 sev_es_rd_ghcb_msr(void)
 
 static __always_inline void sev_es_wr_ghcb_msr(u64 val)
 {
-	u32 low, high;
-
-	low  = (u32)(val);
-	high = (u32)(val >> 32);
-
-	native_wrmsr(MSR_AMD64_SEV_ES_GHCB, low, high);
+	native_wrmsrq(MSR_AMD64_SEV_ES_GHCB, val);
 }
 
 static int vc_fetch_insn_kernel(struct es_em_ctxt *ctxt,
Signature · 1 line
2.49.0
#14 lore raw

Re: [PATCH v1 3/3] x86/msr: Convert a native_wrmsr() use to native_wrmsrq()

Quoted 13 lines · Xin Li (Intel)

Convert a native_wrmsr() use to native_wrmsrq() to zap meaningless type conversions when a u64 MSR value is splitted into two u32.

---
 arch/x86/coco/sev/core.c | 7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)
arch/x86/coco/sev/core.c +0 -0
index ff82151f7718..b3ce6fc8b62d 100644

arch/x86/coco/sev/core.c static inline u64 sev_es_rd_ghcb_msr(void) · line 285

Quoted 10 lines
arch/x86/coco/sev/core.c +1 -6
@@ -282,12 +282,7 @@ static inline u64 sev_es_rd_ghcb_msr(void)
 
 static __always_inline void sev_es_wr_ghcb_msr(u64 val)
 {
-	u32 low, high;
-
-	low  = (u32)(val);
-	high = (u32)(val >> 32);
-
-	native_wrmsr(MSR_AMD64_SEV_ES_GHCB, low, high);
+	native_wrmsrq(MSR_AMD64_SEV_ES_GHCB, val);

BTW., at this point we should probably just replace sev_es_wr_ghcb_msr() calls with direct calls to:

	native_wrmsrq(MSR_AMD64_SEV_ES_GHCB, ...);

as sev_es_wr_ghcb_msr() is now basically an open-coded native_wrmsrq().

Thanks,

Ingo

author
#15 lore raw

Re: [PATCH v1 3/3] x86/msr: Convert a native_wrmsr() use to native_wrmsrq()

Quoted 7 lines · Ingo Molnar
Quoted 3 lines · Xin Li (Intel)

Convert a native_wrmsr() use to native_wrmsrq() to zap meaningless type conversions when a u64 MSR value is splitted into two u32.

Quoted 7 lines

BTW., at this point we should probably just replace sev_es_wr_ghcb_msr() calls with direct calls to:

	native_wrmsrq(MSR_AMD64_SEV_ES_GHCB, ...);

as sev_es_wr_ghcb_msr() is now basically an open-coded native_wrmsrq().

I thought about it, however it looks to me that current code prefers not
to spread MSR_AMD64_SEV_ES_GHCB in 17 callsites. And anyway it's a __always_inline function.

But as you have asked, I will make the change unless someone objects.

Thanks!
Xin

author
#16 lore raw

Re: [PATCH v1 3/3] x86/msr: Convert a native_wrmsr() use to native_wrmsrq()

Quoted 13 lines · Xin Li
Quoted 12 lines · Ingo Molnar
Quoted 3 lines · Xin Li (Intel)

Convert a native_wrmsr() use to native_wrmsrq() to zap meaningless type conversions when a u64 MSR value is splitted into two u32.

BTW., at this point we should probably just replace sev_es_wr_ghcb_msr() calls with direct calls to:

    native_wrmsrq(MSR_AMD64_SEV_ES_GHCB, ...);

Quoted 7 lines
Quoted 2 lines

as sev_es_wr_ghcb_msr() is now basically an open-coded native_wrmsrq().

I thought about it, however it looks to me that current code prefers not
to spread MSR_AMD64_SEV_ES_GHCB in 17 callsites.  And anyway it's a __always_inline function.

But as you have asked, I will make the change unless someone objects.

Hi Ingo,

I took a further look and found that we can't simply replace sev_es_wr_ghcb_msr() with native_wrmsrq(MSR_AMD64_SEV_ES_GHCB, ...).

There are two sev_es_wr_ghcb_msr() definitions. One is defined in arch/x86/boot/compressed/sev.h and it references boot_wrmsr() defined in arch/x86/boot/msr.h to do MSR write.

The other one is defined in arch/x86/include/asm/sev-internal.h, which uses native_wrmsrq() from arch/x86/include/asm/msr.h to write MSR.

Because:
1) arch/x86/boot/startup/sev-shared.c is included in both arch/x86/boot/compressed/sev.c
and
arch/x86/boot/startup/sev-startup.c

2) arch/x86/boot/startup/sev-shared.c has several references to sev_es_wr_ghcb_msr(),

sev_es_wr_ghcb_msr() is converted to boot_wrmsr() when included in arch/x86/boot/compressed/sev.c or native_wrmsrq() when included in arch/x86/boot/startup/sev-startup.c.

It would change the compressed code to use native_wrmsrq() if we remove
sev_es_wr_ghcb_msr() from arch/x86/include/asm/sev-internal.h and use native_wrmsrq() directly in the startup code.

We probably should get rid of boot_wrmsr() and use native_wrmsrq() in the compressed code because they are indeed the same thing. But as we are so close to the v6.16 merge window, I don't think it's a good idea to make the change right now.

So maybe I should just drop this patch and we can do the job after the coming merge window.

But if you think it's not a bad idea to replace native_wrmsr() with native_wrmsrq() right now, I can keep this original patch.

Thanks!
Xin

#17 lore raw

Re: [PATCH v1 3/3] x86/msr: Convert a native_wrmsr() use to native_wrmsrq()

Quoted 23 lines · Xin Li
Quoted 13 lines · Xin Li
Quoted 12 lines · Ingo Molnar
Quoted 3 lines · Xin Li (Intel)

Convert a native_wrmsr() use to native_wrmsrq() to zap meaningless type conversions when a u64 MSR value is splitted into two u32.

BTW., at this point we should probably just replace sev_es_wr_ghcb_msr() calls with direct calls to:

    native_wrmsrq(MSR_AMD64_SEV_ES_GHCB, ...);

Quoted 8 lines
Quoted 2 lines

as sev_es_wr_ghcb_msr() is now basically an open-coded native_wrmsrq().

I thought about it, however it looks to me that current code prefers not to spread MSR_AMD64_SEV_ES_GHCB in 17 callsites.  And anyway it's a __always_inline function.

But as you have asked, I will make the change unless someone objects.

Quoted 8 lines

Hi Ingo,

I took a further look and found that we can't simply replace sev_es_wr_ghcb_msr() with native_wrmsrq(MSR_AMD64_SEV_ES_GHCB, ...).

There are two sev_es_wr_ghcb_msr() definitions. One is defined in arch/x86/boot/compressed/sev.h and it references boot_wrmsr() defined in arch/x86/boot/msr.h to do MSR write.

Ah, indeed, it's also a startup code wrapper, which wrmsrq() doesn't have at the moment. Fair enough.

Thanks,

Ingo

author
#18 lore raw

Re: [PATCH v1 3/3] x86/msr: Convert a native_wrmsr() use to native_wrmsrq()

Quoted 29 lines
Quoted 29 lines
Quoted 15 lines · Xin Li
Quoted 12 lines · Ingo Molnar
Quoted 3 lines · Xin Li (Intel)

Convert a native_wrmsr() use to native_wrmsrq() to zap meaningless type conversions when a u64 MSR value is splitted into two u32.

BTW., at this point we should probably just replace sev_es_wr_ghcb_msr() calls with direct calls to:

    native_wrmsrq(MSR_AMD64_SEV_ES_GHCB, ...);
Quoted 8 lines
Quoted 2 lines

as sev_es_wr_ghcb_msr() is now basically an open-coded native_wrmsrq().

I thought about it, however it looks to me that current code prefers not to spread MSR_AMD64_SEV_ES_GHCB in 17 callsites. And anyway it's a __always_inline function.

But as you have asked, I will make the change unless someone objects.

Hi Ingo,

I took a further look and found that we can't simply replace sev_es_wr_ghcb_msr() with native_wrmsrq(MSR_AMD64_SEV_ES_GHCB, ...).

Quoted 6 lines
Quoted 3 lines

There are two sev_es_wr_ghcb_msr() definitions. One is defined in arch/x86/boot/compressed/sev.h and it references boot_wrmsr() defined in arch/x86/boot/msr.h to do MSR write.

Ah, indeed, it's also a startup code wrapper, which wrmsrq() doesn't have at the moment. Fair enough.

So you want me to drop this patch then?

Thanks!
Xin

#19 lore raw

Re: [PATCH v1 3/3] x86/msr: Convert a native_wrmsr() use to native_wrmsrq()

Quoted 29 lines · Xin Li
Quoted 29 lines
Quoted 29 lines
Quoted 15 lines · Xin Li
Quoted 12 lines · Ingo Molnar
Quoted 3 lines · Xin Li (Intel)

Convert a native_wrmsr() use to native_wrmsrq() to zap meaningless type conversions when a u64 MSR value is splitted into two u32.

BTW., at this point we should probably just replace sev_es_wr_ghcb_msr() calls with direct calls to:

    native_wrmsrq(MSR_AMD64_SEV_ES_GHCB, ...);
Quoted 8 lines
Quoted 2 lines

as sev_es_wr_ghcb_msr() is now basically an open-coded native_wrmsrq().

I thought about it, however it looks to me that current code prefers not to spread MSR_AMD64_SEV_ES_GHCB in 17 callsites. And anyway it's a __always_inline function.

But as you have asked, I will make the change unless someone objects.

Hi Ingo,

I took a further look and found that we can't simply replace sev_es_wr_ghcb_msr() with native_wrmsrq(MSR_AMD64_SEV_ES_GHCB, ...).

Quoted 8 lines
Quoted 6 lines
Quoted 3 lines

There are two sev_es_wr_ghcb_msr() definitions. One is defined in arch/x86/boot/compressed/sev.h and it references boot_wrmsr() defined in arch/x86/boot/msr.h to do MSR write.

Ah, indeed, it's also a startup code wrapper, which wrmsrq() doesn't have at the moment. Fair enough.

So you want me to drop this patch then?

No, patch #3 is fine as-is in its -v1 form, I was wrong.

Thanks,

Ingo

author
#20 lore raw

Re: [PATCH v1 3/3] x86/msr: Convert a native_wrmsr() use to native_wrmsrq()

Quoted 4 lines · Ingo Molnar
Quoted 3 lines
Quoted 2 lines

Ah, indeed, it's also a startup code wrapper, which wrmsrq() doesn't have at the moment. Fair enough.

So you want me to drop this patch then?

No, patch #3 is fine as-is in its -v1 form

Thanks for confirming.

I'll just update patch #2 as version v1A then.

Xin

author
#21 lore raw

Re: [PATCH v1 3/3] x86/msr: Convert a native_wrmsr() use to native_wrmsrq()

In reply to Xin Li (Intel)
Quoted 16 lines · Xin Li (Intel)

Convert a native_wrmsr() use to native_wrmsrq() to zap meaningless type conversions when a u64 MSR value is splitted into two u32.

---
  arch/x86/coco/sev/core.c | 7 +------
  1 file changed, 1 insertion(+), 6 deletions(-)
arch/x86/coco/sev/core.c +0 -0
index ff82151f7718..b3ce6fc8b62d 100644
@@ -282,12 +282,7 @@ static inline u64 sev_es_rd_ghcb_msr(void)
 
  static __always_inline void sev_es_wr_ghcb_msr(u64 val)
  {

arch/x86/coco/sev/core.c static inline u64 sev_es_rd_ghcb_msr(void) · line 288

Quoted 10 lines
arch/x86/coco/sev/core.c +1 -6
-	u32 low, high;
-
-	low  = (u32)(val);
-	high = (u32)(val >> 32);
-
-	native_wrmsr(MSR_AMD64_SEV_ES_GHCB, low, high);
+	native_wrmsrq(MSR_AMD64_SEV_ES_GHCB, val);
  }
 
  static int vc_fetch_insn_kernel(struct es_em_ctxt *ctxt,

Just noticed that this patch doesn't apply to tip/x86/core, I will send it as a separate one.

Thanks!
Xin