* [Qemu-devel] [PATCH RFC v1] i386: hvf: Fix smp boot hangs
@ 2019-01-23 7:34 Heiher
2019-01-23 9:45 ` Paolo Bonzini
0 siblings, 1 reply; 3+ messages in thread
From: Heiher @ 2019-01-23 7:34 UTC (permalink / raw)
To: Paolo Bonzini, Richard Henderson, Eduardo Habkost; +Cc: qemu-devel, Heiher
The machine that with hvf accelerator and smp sometimes boot hangs
because all processors are executing instructions at startup,
including early I/O emulations. We should just allow the bootstrap
processor to initialize the machine and then to wake up slave
processors by interrupt.
Signed-off-by: Heiher <r@hev.cc>
---
target/i386/hvf/hvf.c | 5 -----
1 file changed, 5 deletions(-)
diff --git a/target/i386/hvf/hvf.c b/target/i386/hvf/hvf.c
index e193022c03..0921fcba45 100644
--- a/target/i386/hvf/hvf.c
+++ b/target/i386/hvf/hvf.c
@@ -499,7 +499,6 @@ void hvf_reset_vcpu(CPUState *cpu) {
}
hv_vm_sync_tsc(0);
- cpu->halted = 0;
hv_vcpu_invalidate_tlb(cpu->hvf_fd);
hv_vcpu_flush(cpu->hvf_fd);
}
@@ -582,8 +581,6 @@ int hvf_init_vcpu(CPUState *cpu)
wvmcs(cpu->hvf_fd, VMCS_TPR_THRESHOLD, 0);
- hvf_reset_vcpu(cpu);
-
x86cpu = X86_CPU(cpu);
x86cpu->env.xsave_buf = qemu_memalign(4096, 4096);
@@ -659,8 +656,6 @@ int hvf_vcpu_exec(CPUState *cpu)
int ret = 0;
uint64_t rip = 0;
- cpu->halted = 0;
-
if (hvf_process_events(cpu)) {
return EXCP_HLT;
}
--
2.20.1
^ permalink raw reply related [flat|nested] 3+ messages in thread* Re: [Qemu-devel] [PATCH RFC v1] i386: hvf: Fix smp boot hangs
2019-01-23 7:34 [Qemu-devel] [PATCH RFC v1] i386: hvf: Fix smp boot hangs Heiher
@ 2019-01-23 9:45 ` Paolo Bonzini
2019-01-23 11:05 ` Heiher
0 siblings, 1 reply; 3+ messages in thread
From: Paolo Bonzini @ 2019-01-23 9:45 UTC (permalink / raw)
To: Heiher, Richard Henderson, Eduardo Habkost; +Cc: qemu-devel
On 23/01/19 08:34, Heiher wrote:
> The machine that with hvf accelerator and smp sometimes boot hangs
> because all processors are executing instructions at startup,
> including early I/O emulations. We should just allow the bootstrap
> processor to initialize the machine and then to wake up slave
> processors by interrupt.
>
> Signed-off-by: Heiher <r@hev.cc>
> ---
> target/i386/hvf/hvf.c | 5 -----
> 1 file changed, 5 deletions(-)
>
> diff --git a/target/i386/hvf/hvf.c b/target/i386/hvf/hvf.c
> index e193022c03..0921fcba45 100644
> --- a/target/i386/hvf/hvf.c
> +++ b/target/i386/hvf/hvf.c
> @@ -499,7 +499,6 @@ void hvf_reset_vcpu(CPUState *cpu) {
> }
>
> hv_vm_sync_tsc(0);
> - cpu->halted = 0;
> hv_vcpu_invalidate_tlb(cpu->hvf_fd);
> hv_vcpu_flush(cpu->hvf_fd);
> }
> @@ -582,8 +581,6 @@ int hvf_init_vcpu(CPUState *cpu)
>
> wvmcs(cpu->hvf_fd, VMCS_TPR_THRESHOLD, 0);
>
> - hvf_reset_vcpu(cpu);
Why this?
> x86cpu = X86_CPU(cpu);
> x86cpu->env.xsave_buf = qemu_memalign(4096, 4096);
>
> @@ -659,8 +656,6 @@ int hvf_vcpu_exec(CPUState *cpu)
> int ret = 0;
> uint64_t rip = 0;
>
> - cpu->halted = 0;
> -
> if (hvf_process_events(cpu)) {
> return EXCP_HLT;
> }
>
Looks good apart from the question above.
Paoo
^ permalink raw reply [flat|nested] 3+ messages in thread* Re: [Qemu-devel] [PATCH RFC v1] i386: hvf: Fix smp boot hangs
2019-01-23 9:45 ` Paolo Bonzini
@ 2019-01-23 11:05 ` Heiher
0 siblings, 0 replies; 3+ messages in thread
From: Heiher @ 2019-01-23 11:05 UTC (permalink / raw)
To: Paolo Bonzini; +Cc: Richard Henderson, Eduardo Habkost, qemu-devel
On Wed, Jan 23, 2019 at 5:45 PM Paolo Bonzini <pbonzini@redhat.com> wrote:
>
> On 23/01/19 08:34, Heiher wrote:
> > The machine that with hvf accelerator and smp sometimes boot hangs
> > because all processors are executing instructions at startup,
> > including early I/O emulations. We should just allow the bootstrap
> > processor to initialize the machine and then to wake up slave
> > processors by interrupt.
> >
> > Signed-off-by: Heiher <r@hev.cc>
> > ---
> > target/i386/hvf/hvf.c | 5 -----
> > 1 file changed, 5 deletions(-)
> >
> > diff --git a/target/i386/hvf/hvf.c b/target/i386/hvf/hvf.c
> > index e193022c03..0921fcba45 100644
> > --- a/target/i386/hvf/hvf.c
> > +++ b/target/i386/hvf/hvf.c
> > @@ -499,7 +499,6 @@ void hvf_reset_vcpu(CPUState *cpu) {
> > }
> >
> > hv_vm_sync_tsc(0);
> > - cpu->halted = 0;
> > hv_vcpu_invalidate_tlb(cpu->hvf_fd);
> > hv_vcpu_flush(cpu->hvf_fd);
> > }
> > @@ -582,8 +581,6 @@ int hvf_init_vcpu(CPUState *cpu)
> >
> > wvmcs(cpu->hvf_fd, VMCS_TPR_THRESHOLD, 0);
> >
> > - hvf_reset_vcpu(cpu);
>
> Why this?
It looks redundant. I think it is enough to reset vcpu from
x86_cpu_reset, likes kvm.
https://git.qemu.org/?p=qemu.git;a=blob;f=target/i386/cpu.c;h=2f5412592d30ea7a24810f3bf6a8c7d4b035a0f9;hb=952bc8b3c2cbba78261923a1e8ca55cda261dee9#l4732
Do you think it is correct?
>
> > x86cpu = X86_CPU(cpu);
> > x86cpu->env.xsave_buf = qemu_memalign(4096, 4096);
> >
> > @@ -659,8 +656,6 @@ int hvf_vcpu_exec(CPUState *cpu)
> > int ret = 0;
> > uint64_t rip = 0;
> >
> > - cpu->halted = 0;
> > -
> > if (hvf_process_events(cpu)) {
> > return EXCP_HLT;
> > }
> >
>
> Looks good apart from the question above.
>
> Paoo
Thank you reply.
--
Best regards!
Hev
https://hev.cc
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2019-01-23 11:06 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2019-01-23 7:34 [Qemu-devel] [PATCH RFC v1] i386: hvf: Fix smp boot hangs Heiher
2019-01-23 9:45 ` Paolo Bonzini
2019-01-23 11:05 ` Heiher
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).