Commit 383c2525 authored by Sebastian Andrzej Siewior's avatar Sebastian Andrzej Siewior Committed by Borislav Petkov

x86/entry: Add TIF_NEED_FPU_LOAD

Add TIF_NEED_FPU_LOAD. This flag is used for loading the FPU registers
before returning to userland. It must not be set on systems without a
FPU.

If this flag is cleared, the CPU's FPU registers hold the latest,
up-to-date content of the current task's (current()) FPU registers.
The in-memory copy (union fpregs_state) is not valid.

If this flag is set, then all of CPU's FPU registers may hold a random
value (except for PKRU) and it is required to load the content of the
FPU registers on return to userland.

Introduce it now as a preparatory change before adding the main feature.
Signed-off-by: default avatarSebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: default avatarBorislav Petkov <bp@suse.de>
Reviewed-by: default avatarDave Hansen <dave.hansen@intel.com>
Reviewed-by: default avatarThomas Gleixner <tglx@linutronix.de>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Aubrey Li <aubrey.li@intel.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jann Horn <jannh@google.com>
Cc: "Jason A. Donenfeld" <Jason@zx2c4.com>
Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Cc: kvm ML <kvm@vger.kernel.org>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Radim Krčmář <rkrcmar@redhat.com>
Cc: Rik van Riel <riel@surriel.com>
Cc: Tim Chen <tim.c.chen@linux.intel.com>
Cc: x86-ml <x86@kernel.org>
Link: https://lkml.kernel.org/r/20190403164156.19645-17-bigeasy@linutronix.de
parent 0cecca9d
...@@ -508,6 +508,14 @@ static inline void __fpregs_load_activate(struct fpu *fpu, int cpu) ...@@ -508,6 +508,14 @@ static inline void __fpregs_load_activate(struct fpu *fpu, int cpu)
* - switch_fpu_finish() restores the new state as * - switch_fpu_finish() restores the new state as
* necessary. * necessary.
* *
* If TIF_NEED_FPU_LOAD is cleared then the CPU's FPU registers
* are saved in the current thread's FPU register state.
*
* If TIF_NEED_FPU_LOAD is set then CPU's FPU registers may not
* hold current()'s FPU registers. It is required to load the
* registers before returning to userland or using the content
* otherwise.
*
* The FPU context is only stored/restored for a user task and * The FPU context is only stored/restored for a user task and
* ->mm is used to distinguish between kernel and user threads. * ->mm is used to distinguish between kernel and user threads.
*/ */
......
...@@ -88,6 +88,7 @@ struct thread_info { ...@@ -88,6 +88,7 @@ struct thread_info {
#define TIF_USER_RETURN_NOTIFY 11 /* notify kernel of userspace return */ #define TIF_USER_RETURN_NOTIFY 11 /* notify kernel of userspace return */
#define TIF_UPROBE 12 /* breakpointed or singlestepping */ #define TIF_UPROBE 12 /* breakpointed or singlestepping */
#define TIF_PATCH_PENDING 13 /* pending live patching update */ #define TIF_PATCH_PENDING 13 /* pending live patching update */
#define TIF_NEED_FPU_LOAD 14 /* load FPU on return to userspace */
#define TIF_NOCPUID 15 /* CPUID is not accessible in userland */ #define TIF_NOCPUID 15 /* CPUID is not accessible in userland */
#define TIF_NOTSC 16 /* TSC is not accessible in userland */ #define TIF_NOTSC 16 /* TSC is not accessible in userland */
#define TIF_IA32 17 /* IA32 compatibility process */ #define TIF_IA32 17 /* IA32 compatibility process */
...@@ -117,6 +118,7 @@ struct thread_info { ...@@ -117,6 +118,7 @@ struct thread_info {
#define _TIF_USER_RETURN_NOTIFY (1 << TIF_USER_RETURN_NOTIFY) #define _TIF_USER_RETURN_NOTIFY (1 << TIF_USER_RETURN_NOTIFY)
#define _TIF_UPROBE (1 << TIF_UPROBE) #define _TIF_UPROBE (1 << TIF_UPROBE)
#define _TIF_PATCH_PENDING (1 << TIF_PATCH_PENDING) #define _TIF_PATCH_PENDING (1 << TIF_PATCH_PENDING)
#define _TIF_NEED_FPU_LOAD (1 << TIF_NEED_FPU_LOAD)
#define _TIF_NOCPUID (1 << TIF_NOCPUID) #define _TIF_NOCPUID (1 << TIF_NOCPUID)
#define _TIF_NOTSC (1 << TIF_NOTSC) #define _TIF_NOTSC (1 << TIF_NOTSC)
#define _TIF_IA32 (1 << TIF_IA32) #define _TIF_IA32 (1 << TIF_IA32)
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment