Message ID | 20231122025811.2973-1-npiggin@gmail.com (mailing list archive) |
---|---|
State | Accepted |
Commit | dc158d23b33df9033bcc8e7117e8591dd2f9d125 |
Headers | show |
Series | KVM: PPC: Book3S HV: Fix KVM_RUN clobbering FP/VEC user registers | expand |
Context | Check | Description |
---|---|---|
snowpatch_ozlabs/github-powerpc_ppctests | success | Successfully ran 8 jobs. |
snowpatch_ozlabs/github-powerpc_selftests | success | Successfully ran 8 jobs. |
snowpatch_ozlabs/github-powerpc_sparse | success | Successfully ran 4 jobs. |
snowpatch_ozlabs/github-powerpc_clang | success | Successfully ran 6 jobs. |
On Wed, 22 Nov 2023 12:58:11 +1000, Nicholas Piggin wrote: > Before running a guest, the host process (e.g., QEMU) FP/VEC registers > are saved if they were being used, similarly to when the kernel uses FP > registers. The guest values are then be loaded into regs, and the host > process registers will be restored lazily when it uses FP/VEC. > > KVM HV has a bug here: the host process registers do get saved, but the > user MSR bits remain enabled, which indicates the registers are valid > for the process. After they are clobbered by running the guest, this > valid indication causes the host process to take on the FP/VEC regiter > values of the guest. > > [...] Applied to powerpc/fixes. [1/1] KVM: PPC: Book3S HV: Fix KVM_RUN clobbering FP/VEC user registers https://git.kernel.org/powerpc/c/dc158d23b33df9033bcc8e7117e8591dd2f9d125 cheers
diff --git a/arch/powerpc/kernel/process.c b/arch/powerpc/kernel/process.c index 392404688cec..9452a54d356c 100644 --- a/arch/powerpc/kernel/process.c +++ b/arch/powerpc/kernel/process.c @@ -1198,11 +1198,11 @@ void kvmppc_save_user_regs(void) usermsr = current->thread.regs->msr; + /* Caller has enabled FP/VEC/VSX/TM in MSR */ if (usermsr & MSR_FP) - save_fpu(current); - + __giveup_fpu(current); if (usermsr & MSR_VEC) - save_altivec(current); + __giveup_altivec(current); #ifdef CONFIG_PPC_TRANSACTIONAL_MEM if (usermsr & MSR_TM) {
Before running a guest, the host process (e.g., QEMU) FP/VEC registers are saved if they were being used, similarly to when the kernel uses FP registers. The guest values are then be loaded into regs, and the host process registers will be restored lazily when it uses FP/VEC. KVM HV has a bug here: the host process registers do get saved, but the user MSR bits remain enabled, which indicates the registers are valid for the process. After they are clobbered by running the guest, this valid indication causes the host process to take on the FP/VEC regiter values of the guest. Fixes: de2a20aa7237b ("powerpc: Prepare for splitting giveup_{fpu, altivec, vsx} in two") Signed-off-by: Nicholas Piggin <npiggin@gmail.com> --- arch/powerpc/kernel/process.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)