x86/fpu: minor cleanups

PATCH Applied Oleg Nesterov 4 participants 12 messages Mar 2, 2015 – Mar 10, 2015 4 messages not in the archive x86

Share & tools

A message that is not in the archive (sent off-list or not yet received): CAJ_zFkJkHpU4Y8wrrZjO0VnhzUVqJfyrLmNfRsBRT29bgbDSHg@mail.gmail.com

A message that is not in the archive (sent off-list or not yet received): [email protected]

Oleg Nesterov

author

2015-03-02 18:34 UTC #4 lore raw

cover x86/fpu: minor cleanups

add lkml/cc's.

Quoted 8 lines · Oleg Nesterov

Btw, why xstateregs_get/set() looks so confusing? The comment says "Copy the 48bytes defined by the software first into the xstate" but the code uses ->fxsave. Yes, this is the same memory, but still. I'll send the cosmetic cleanup.
Or fx_finit()... Again, memset(fx, 0, xstate_size) is correct but only because of the current layout. I think this needs a cleanup too.

Of course this is purely cosmetic, but still...

Oleg.

Oleg Nesterov

author

2015-03-02 18:34 UTC #5 lore raw

PATCH 1/2 x86/regsets: change xstateregs_get/set to use ->xsave.i387 rather than ->fxsave

Cosmetic. xstateregs_get() and xstateregs_set() abuse ->fxsave to access xsave->i387.sw_reserved. This is correct, ->fxsave and xsave->i387 share the same memory, but imho this looks confusing.

And we can make this code more readable if we add "struct xsave_struct *" local.

Signed-off-by Oleg Nesterov <[email protected]>

---
 arch/x86/kernel/i387.c |   25 +++++++++----------------
 1 files changed, 9 insertions(+), 16 deletions(-)

arch/x86/kernel/i387.c +9 ~~-16~~

index 8e070a6..4b12df8 100644
@@ -339,6 +339,7 @@ int xstateregs_get(struct task_struct *target, const struct user_regset *regset,
 		unsigned int pos, unsigned int count,
 		void *kbuf, void __user *ubuf)
 {
+	struct xsave_struct *xsave = &target->thread.fpu.state->xsave;
 	int ret;
 
 	if (!cpu_has_xsave)

@@ -353,14 +354,12 @@ int xstateregs_get(struct task_struct *target, const struct user_regset *regset,
 	 * memory layout in the thread struct, so that we can copy the entire
 	 * xstateregs to the user using one user_regset_copyout().
 	 */
-	memcpy(&target->thread.fpu.state->fxsave.sw_reserved,
-	       xstate_fx_sw_bytes, sizeof(xstate_fx_sw_bytes));
-
+	memcpy(&xsave->i387.sw_reserved,
+		xstate_fx_sw_bytes, sizeof(xstate_fx_sw_bytes));
 	/*
 	 * Copy the xstate memory layout.
 	 */
-	ret = user_regset_copyout(&pos, &count, &kbuf, &ubuf,
-				  &target->thread.fpu.state->xsave, 0, -1);
+	ret = user_regset_copyout(&pos, &count, &kbuf, &ubuf, xsave, 0, -1);
 	return ret;
 }

@@ -368,8 +367,8 @@ int xstateregs_set(struct task_struct *target, const struct user_regset *regset,
 		  unsigned int pos, unsigned int count,
 		  const void *kbuf, const void __user *ubuf)
 {
+	struct xsave_struct *xsave = &target->thread.fpu.state->xsave;
 	int ret;
-	struct xsave_hdr_struct *xsave_hdr;
 
 	if (!cpu_has_xsave)
 		return -ENODEV;

@@ -378,22 +377,16 @@ int xstateregs_set(struct task_struct *target, const struct user_regset *regset,
 	if (ret)
 		return ret;
 
-	ret = user_regset_copyin(&pos, &count, &kbuf, &ubuf,
-				 &target->thread.fpu.state->xsave, 0, -1);
-
+	ret = user_regset_copyin(&pos, &count, &kbuf, &ubuf, xsave, 0, -1);
 	/*
 	 * mxcsr reserved bits must be masked to zero for security reasons.
 	 */
-	target->thread.fpu.state->fxsave.mxcsr &= mxcsr_feature_mask;
-
-	xsave_hdr = &target->thread.fpu.state->xsave.xsave_hdr;
-
-	xsave_hdr->xstate_bv &= pcntxt_mask;
+	xsave->i387.mxcsr &= mxcsr_feature_mask;
+	xsave->xsave_hdr.xstate_bv &= pcntxt_mask;
 	/*
 	 * These bits must be zero.
 	 */
-	memset(xsave_hdr->reserved, 0, 48);
-
+	memset(&xsave->xsave_hdr.reserved, 0, 48);
 	return ret;
 }

Signature · 1 line

1.5.5.1

Rik van Riel

Reviewed-by on v1 1/2

2015-03-03 19:23 UTC #6 lore raw

Re: [PATCH 1/2] x86/regsets: change xstateregs_get/set to use ->xsave.i387 rather than ->fxsave

Quoted 8 lines · Oleg Nesterov

Cosmetic. xstateregs_get() and xstateregs_set() abuse ->fxsave to access xsave->i387.sw_reserved. This is correct, ->fxsave and xsave->i387 share the same memory, but imho this looks confusing.
And we can make this code more readable if we add "struct xsave_struct *" local.
Signed-off-by Oleg Nesterov <[email protected]>

Reviewed-by Rik van Riel <[email protected]>

Signature · 1 line

All rights reversed

Borislav Petkov

maintainer Applied

2015-03-04 10:31 UTC #7 lore raw

Re: [PATCH 1/2] x86/regsets: change xstateregs_get/set to use ->xsave.i387 rather than ->fxsave

In reply to Oleg Nesterov

Quoted 8 lines · Oleg Nesterov

Cosmetic. xstateregs_get() and xstateregs_set() abuse ->fxsave to access xsave->i387.sw_reserved. This is correct, ->fxsave and xsave->i387 share the same memory, but imho this looks confusing.
And we can make this code more readable if we add "struct xsave_struct *" local.
Signed-off-by Oleg Nesterov <[email protected]>

Applied, thanks.

Signature · 5 lines

Regards/Gruss,
    Boris.

ECO tip #101: Trim your mails when you reply.
--

tip-bot for Oleg Nesterov

merge bot

2015-03-10 10:08 UTC #8 lore raw

[tip:x86/fpu] x86/fpu: Change xstateregs_get()/set() to use -> xsave.i387 rather than ->fxsave

In reply to Oleg Nesterov

x86/fpu: Change xstateregs_get()/set() to use ->xsave.i387 rather than ->fxsave

Commit-ID:  e7f180dcd8ab48f18b20d7e8a7e9b39192bdf8e0
Gitweb:     http://git.kernel.org/tip/e7f180dcd8ab48f18b20d7e8a7e9b39192bdf8e0
Author:     Oleg Nesterov <[email protected]>
AuthorDate: Tue, 10 Mar 2015 07:06:24 +0100
Committer:  Ingo Molnar <[email protected]>
CommitDate: Tue, 10 Mar 2015 07:14:31 +0100

x86/fpu: Change xstateregs_get()/set() to use ->xsave.i387 rather than ->fxsave

This is a cosmetic change: xstateregs_get() and xstateregs_set() abuse ->fxsave to access xsave->i387.sw_reserved.

This practice is correct, ->fxsave and xsave->i387 share the same memory, but IMHO this looks confusing.

And we can make this code more readable if we add a "struct xsave_struct *" local variable as well.

Signed-off-by Oleg Nesterov <[email protected]>

Signed-off-by Borislav Petkov <[email protected]>

Reviewed-by Rik van Riel <[email protected]>

Cc Andy Lutomirski <[email protected]>

Cc Borislav Petkov <[email protected]>

Cc H. Peter Anvin <[email protected]>

Cc Linus Torvalds <[email protected]>

Cc Tavis Ormandy <[email protected]>

Cc Thomas Gleixner <[email protected]>

Link http://lkml.kernel.org/r/[email protected] open here

Signed-off-by Ingo Molnar <[email protected]>

---
 arch/x86/kernel/i387.c | 25 +++++++++----------------
 1 file changed, 9 insertions(+), 16 deletions(-)

arch/x86/kernel/i387.c +9 ~~-16~~

index 8416b5f..03cc0ad 100644
@@ -339,6 +339,7 @@ int xstateregs_get(struct task_struct *target, const struct user_regset *regset,
 		unsigned int pos, unsigned int count,
 		void *kbuf, void __user *ubuf)
 {
+	struct xsave_struct *xsave = &target->thread.fpu.state->xsave;
 	int ret;
 
 	if (!cpu_has_xsave)

@@ -353,14 +354,12 @@ int xstateregs_get(struct task_struct *target, const struct user_regset *regset,
 	 * memory layout in the thread struct, so that we can copy the entire
 	 * xstateregs to the user using one user_regset_copyout().
 	 */
-	memcpy(&target->thread.fpu.state->fxsave.sw_reserved,
-	       xstate_fx_sw_bytes, sizeof(xstate_fx_sw_bytes));
-
+	memcpy(&xsave->i387.sw_reserved,
+		xstate_fx_sw_bytes, sizeof(xstate_fx_sw_bytes));
 	/*
 	 * Copy the xstate memory layout.
 	 */
-	ret = user_regset_copyout(&pos, &count, &kbuf, &ubuf,
-				  &target->thread.fpu.state->xsave, 0, -1);
+	ret = user_regset_copyout(&pos, &count, &kbuf, &ubuf, xsave, 0, -1);
 	return ret;
 }

@@ -368,8 +367,8 @@ int xstateregs_set(struct task_struct *target, const struct user_regset *regset,
 		  unsigned int pos, unsigned int count,
 		  const void *kbuf, const void __user *ubuf)
 {
+	struct xsave_struct *xsave = &target->thread.fpu.state->xsave;
 	int ret;
-	struct xsave_hdr_struct *xsave_hdr;
 
 	if (!cpu_has_xsave)
 		return -ENODEV;

@@ -378,22 +377,16 @@ int xstateregs_set(struct task_struct *target, const struct user_regset *regset,
 	if (ret)
 		return ret;
 
-	ret = user_regset_copyin(&pos, &count, &kbuf, &ubuf,
-				 &target->thread.fpu.state->xsave, 0, -1);
-
+	ret = user_regset_copyin(&pos, &count, &kbuf, &ubuf, xsave, 0, -1);
 	/*
 	 * mxcsr reserved bits must be masked to zero for security reasons.
 	 */
-	target->thread.fpu.state->fxsave.mxcsr &= mxcsr_feature_mask;
-
-	xsave_hdr = &target->thread.fpu.state->xsave.xsave_hdr;
-
-	xsave_hdr->xstate_bv &= pcntxt_mask;
+	xsave->i387.mxcsr &= mxcsr_feature_mask;
+	xsave->xsave_hdr.xstate_bv &= pcntxt_mask;
 	/*
 	 * These bits must be zero.
 	 */
-	memset(xsave_hdr->reserved, 0, 48);
-
+	memset(&xsave->xsave_hdr.reserved, 0, 48);
 	return ret;
 }

Oleg Nesterov

author

2015-03-02 18:34 UTC #9 lore raw

PATCH 2/2 x86/fpu: factor out memset(xstate, 0) in fpu_finit() paths

In reply to Oleg Nesterov

fx_finit() has 2 users but only fpu_finit() needs to nullify xstate, alloc_bootmem_align() in setup_init_fpu_buf() returns zero-filled memory.

And note that both memset()'s look confusing. Yes, offsetof() is 0 for ->fxsave or ->fsave, but it would be more clean to turn them into a single memset() which nullifies fpu->state.

Signed-off-by Oleg Nesterov <[email protected]>

---
 arch/x86/include/asm/fpu-internal.h |    1 -
 arch/x86/kernel/i387.c              |    3 ++-
 2 files changed, 2 insertions(+), 2 deletions(-)

arch/x86/include/asm/fpu-internal.h +0 -1

index 02f2e08..8809b5a 100644
@@ -107,7 +107,6 @@ static __always_inline __pure bool use_fxsr(void)
 
 static inline void fx_finit(struct i387_fxsave_struct *fx)
 {
-	memset(fx, 0, xstate_size);
 	fx->cwd = 0x37f;
 	fx->mxcsr = MXCSR_DEFAULT;
 }

arch/x86/kernel/i387.c +2 -1

index 4b12df8..9b7759f 100644
@@ -224,11 +224,12 @@ void fpu_finit(struct fpu *fpu)
 		return;
 	}
 
+	memset(fpu->state, 0, xstate_size);
+
 	if (cpu_has_fxsr) {
 		fx_finit(&fpu->state->fxsave);
 	} else {
 		struct i387_fsave_struct *fp = &fpu->state->fsave;
-		memset(fp, 0, xstate_size);
 		fp->cwd = 0xffff037fu;
 		fp->swd = 0xffff0000u;
 		fp->twd = 0xffffffffu;

Signature · 1 line

1.5.5.1

Rik van Riel

Acked-by on v1 2/2

2015-03-03 22:01 UTC #10 lore raw

Re: [PATCH 2/2] x86/fpu: factor out memset(xstate, 0) in fpu_finit() paths

Quoted 9 lines · Oleg Nesterov

fx_finit() has 2 users but only fpu_finit() needs to nullify xstate, alloc_bootmem_align() in setup_init_fpu_buf() returns zero-filled memory.
And note that both memset()'s look confusing. Yes, offsetof() is 0 for ->fxsave or ->fsave, but it would be more clean to turn them into a single memset() which nullifies fpu->state.
Signed-off-by Oleg Nesterov <[email protected]>

Acked-by Rik van Riel <[email protected]>

Signature · 1 line

All rights reversed

Borislav Petkov

maintainer Applied

2015-03-04 11:39 UTC #11 lore raw

Re: [PATCH 2/2] x86/fpu: factor out memset(xstate, 0) in fpu_finit() paths

In reply to Oleg Nesterov

Quoted 9 lines · Oleg Nesterov

fx_finit() has 2 users but only fpu_finit() needs to nullify xstate, alloc_bootmem_align() in setup_init_fpu_buf() returns zero-filled memory.
And note that both memset()'s look confusing. Yes, offsetof() is 0 for ->fxsave or ->fsave, but it would be more clean to turn them into a single memset() which nullifies fpu->state.
Signed-off-by Oleg Nesterov <[email protected]>

Applied, thanks.

Signature · 5 lines

Regards/Gruss,
    Boris.

ECO tip #101: Trim your mails when you reply.
--

tip-bot for Oleg Nesterov

merge bot

2015-03-10 10:08 UTC #12 lore raw

[tip:x86/fpu] x86/fpu: Factor out memset(xstate, 0) in fpu_finit( ) paths

In reply to Oleg Nesterov

x86/fpu: Factor out memset(xstate, 0) in fpu_finit() paths

Commit-ID:  1d23c4518b1f3a03c278f23333149245c178d2a6
Gitweb:     http://git.kernel.org/tip/1d23c4518b1f3a03c278f23333149245c178d2a6
Author:     Oleg Nesterov <[email protected]>
AuthorDate: Tue, 10 Mar 2015 07:06:25 +0100
Committer:  Ingo Molnar <[email protected]>
CommitDate: Tue, 10 Mar 2015 07:14:31 +0100

x86/fpu: Factor out memset(xstate, 0) in fpu_finit() paths

fx_finit() has two users but only fpu_finit() needs to clear xstate, alloc_bootmem_align() in setup_init_fpu_buf() returns zero-filled memory.

And note that both memset()'s look confusing. Yes, offsetof() is 0 for ->fxsave or ->fsave, but it would be cleaner to turn them into a single memset() which zeroes fpu->state.

Signed-off-by Oleg Nesterov <[email protected]>

Signed-off-by Borislav Petkov <[email protected]>

Acked-by Rik van Riel <[email protected]>

Cc Andy Lutomirski <[email protected]>

Cc Borislav Petkov <[email protected]>

Cc H. Peter Anvin <[email protected]>

Cc Linus Torvalds <[email protected]>

Cc Tavis Ormandy <[email protected]>

Cc Thomas Gleixner <[email protected]>

Link http://lkml.kernel.org/r/[email protected] open here

Signed-off-by Ingo Molnar <[email protected]>

---
 arch/x86/include/asm/fpu-internal.h | 1 -
 arch/x86/kernel/i387.c              | 3 ++-
 2 files changed, 2 insertions(+), 2 deletions(-)

arch/x86/include/asm/fpu-internal.h +0 -1

index 61609b9..5fa1be2 100644
@@ -135,7 +135,6 @@ static __always_inline __pure bool use_fxsr(void)
 
 static inline void fx_finit(struct i387_fxsave_struct *fx)
 {
-	memset(fx, 0, xstate_size);
 	fx->cwd = 0x37f;
 	fx->mxcsr = MXCSR_DEFAULT;
 }

arch/x86/kernel/i387.c +2 -1

index 03cc0ad..0f3de66 100644
@@ -224,11 +224,12 @@ void fpu_finit(struct fpu *fpu)
 		return;
 	}
 
+	memset(fpu->state, 0, xstate_size);
+
 	if (cpu_has_fxsr) {
 		fx_finit(&fpu->state->fxsave);
 	} else {
 		struct i387_fsave_struct *fp = &fpu->state->fsave;
-		memset(fp, 0, xstate_size);
 		fp->cwd = 0xffff037fu;
 		fp->swd = 0xffff0000u;
 		fp->twd = 0xffffffffu;

Borislav Petkov

maintainer

2015-03-09 18:09 UTC #13 lore raw

[GIT PULL] x86/fpu cleanups

Hi Ingo,

please pull two more cleanups to the FPU insanity from Oleg.

So this is *not* the urgent stuff - the urgent stuff will come later. This is the unload-some-more-onto-tip stuff which can be safely queued for 4.1 and I can forget about it here. :)

Thanks.

---
The following changes since commit ae486033b980346eb6a77240101210cb66924a91:

Merge tag 'tip_x86_fpu' of git://git.kernel.org/pub/scm/linux/kernel/git/bp/bp into x86/fpu (2015-03-03 12:12:15 +0100)

are available in the git repository at:

git://git.kernel.org/pub/scm/linux/kernel/git/bp/bp.git tags/tip_x86_fpu_2

for you to fetch changes up to 9ed010d31f173be786f0398a64f08bb4a0300c3b:

x86/fpu: factor out memset(xstate, 0) in fpu_finit() paths (2015-03-04 11:57:46 +0100)

----------------------------------------------------------------
Two more cleanups from Oleg.

----------------------------------------------------------------
Oleg Nesterov (2):
x86/fpu: Change xstateregs_get()/set() to use ->xsave.i387 rather than ->fxsave x86/fpu: factor out memset(xstate, 0) in fpu_finit() paths

 arch/x86/include/asm/fpu-internal.h |  1 -
 arch/x86/kernel/i387.c              | 28 +++++++++++-----------------
 2 files changed, 11 insertions(+), 18 deletions(-)

Signature · 5 lines

Regards/Gruss,
    Boris.

ECO tip #101: Trim your mails when you reply.
--

Borislav Petkov

maintainer

2015-03-10 06:07 UTC #14 lore raw

PATCH 1/2 x86/fpu: Change xstateregs_get()/set() to use ->xsave.i387 rather than ->fxsave

From: Oleg Nesterov <[email protected]>

This is a cosmetic change: xstateregs_get() and xstateregs_set() abuse ->fxsave to access xsave->i387.sw_reserved. This is correct, ->fxsave and xsave->i387 share the same memory, but IMHO this looks confusing.

And we can make this code more readable if we add "struct xsave_struct *" local.

Signed-off-by Oleg Nesterov <[email protected]>

Reviewed-by Rik van Riel <[email protected]>

Cc Tavis Ormandy <[email protected]>

Cc Andy Lutomirski <[email protected]>

Cc <[email protected]>

Cc Linus Torvalds <[email protected]>

Link http://lkml.kernel.org/r/[email protected] open here

Signed-off-by Borislav Petkov <[email protected]>

---
 arch/x86/kernel/i387.c | 25 +++++++++----------------
 1 file changed, 9 insertions(+), 16 deletions(-)

arch/x86/kernel/i387.c +9 ~~-16~~

index 8416b5f85806..03cc0add8694 100644
@@ -339,6 +339,7 @@ int xstateregs_get(struct task_struct *target, const struct user_regset *regset,
 		unsigned int pos, unsigned int count,
 		void *kbuf, void __user *ubuf)
 {
+	struct xsave_struct *xsave = &target->thread.fpu.state->xsave;
 	int ret;
 
 	if (!cpu_has_xsave)

@@ -353,14 +354,12 @@ int xstateregs_get(struct task_struct *target, const struct user_regset *regset,
 	 * memory layout in the thread struct, so that we can copy the entire
 	 * xstateregs to the user using one user_regset_copyout().
 	 */
-	memcpy(&target->thread.fpu.state->fxsave.sw_reserved,
-	       xstate_fx_sw_bytes, sizeof(xstate_fx_sw_bytes));
-
+	memcpy(&xsave->i387.sw_reserved,
+		xstate_fx_sw_bytes, sizeof(xstate_fx_sw_bytes));
 	/*
 	 * Copy the xstate memory layout.
 	 */
-	ret = user_regset_copyout(&pos, &count, &kbuf, &ubuf,
-				  &target->thread.fpu.state->xsave, 0, -1);
+	ret = user_regset_copyout(&pos, &count, &kbuf, &ubuf, xsave, 0, -1);
 	return ret;
 }

@@ -368,8 +367,8 @@ int xstateregs_set(struct task_struct *target, const struct user_regset *regset,
 		  unsigned int pos, unsigned int count,
 		  const void *kbuf, const void __user *ubuf)
 {
+	struct xsave_struct *xsave = &target->thread.fpu.state->xsave;
 	int ret;
-	struct xsave_hdr_struct *xsave_hdr;
 
 	if (!cpu_has_xsave)
 		return -ENODEV;

@@ -378,22 +377,16 @@ int xstateregs_set(struct task_struct *target, const struct user_regset *regset,
 	if (ret)
 		return ret;
 
-	ret = user_regset_copyin(&pos, &count, &kbuf, &ubuf,
-				 &target->thread.fpu.state->xsave, 0, -1);
-
+	ret = user_regset_copyin(&pos, &count, &kbuf, &ubuf, xsave, 0, -1);
 	/*
 	 * mxcsr reserved bits must be masked to zero for security reasons.
 	 */
-	target->thread.fpu.state->fxsave.mxcsr &= mxcsr_feature_mask;
-
-	xsave_hdr = &target->thread.fpu.state->xsave.xsave_hdr;
-
-	xsave_hdr->xstate_bv &= pcntxt_mask;
+	xsave->i387.mxcsr &= mxcsr_feature_mask;
+	xsave->xsave_hdr.xstate_bv &= pcntxt_mask;
 	/*
 	 * These bits must be zero.
 	 */
-	memset(xsave_hdr->reserved, 0, 48);
-
+	memset(&xsave->xsave_hdr.reserved, 0, 48);
 	return ret;
 }

Signature · 1 line

2.2.0.33.gc18b867

Borislav Petkov

maintainer

2015-03-10 06:07 UTC #15 lore raw

PATCH 2/2 x86/fpu: factor out memset(xstate, 0) in fpu_finit() paths

From: Oleg Nesterov <[email protected]>

fx_finit() has 2 users but only fpu_finit() needs to nullify xstate, alloc_bootmem_align() in setup_init_fpu_buf() returns zero-filled memory.

And note that both memset()'s look confusing. Yes, offsetof() is 0 for ->fxsave or ->fsave, but it would be more clean to turn them into a single memset() which nullifies fpu->state.

Signed-off-by Oleg Nesterov <[email protected]>

Acked-by Rik van Riel <[email protected]>

Cc Andy Lutomirski <[email protected]>

Cc Linus Torvalds <[email protected]>

Cc Tavis Ormandy <[email protected]>

Cc <[email protected]>

Link http://lkml.kernel.org/r/[email protected] open here

Signed-off-by Borislav Petkov <[email protected]>

---
 arch/x86/include/asm/fpu-internal.h | 1 -
 arch/x86/kernel/i387.c              | 3 ++-
 2 files changed, 2 insertions(+), 2 deletions(-)

arch/x86/include/asm/fpu-internal.h +0 -1

index 61609b963eab..5fa1be21ac2a 100644
@@ -135,7 +135,6 @@ static __always_inline __pure bool use_fxsr(void)
 
 static inline void fx_finit(struct i387_fxsave_struct *fx)
 {
-	memset(fx, 0, xstate_size);
 	fx->cwd = 0x37f;
 	fx->mxcsr = MXCSR_DEFAULT;
 }

arch/x86/kernel/i387.c +2 -1

index 03cc0add8694..0f3de6674ae3 100644
@@ -224,11 +224,12 @@ void fpu_finit(struct fpu *fpu)
 		return;
 	}
 
+	memset(fpu->state, 0, xstate_size);
+
 	if (cpu_has_fxsr) {
 		fx_finit(&fpu->state->fxsave);
 	} else {
 		struct i387_fsave_struct *fp = &fpu->state->fsave;
-		memset(fp, 0, xstate_size);
 		fp->cwd = 0xffff037fu;
 		fp->swd = 0xffff0000u;
 		fp->twd = 0xffffffffu;

Signature · 1 line

2.2.0.33.gc18b867