Re: [Linux PPC] Disable PREEMPT

Hi Christian, Michal,

quoted

quoted

quoted

quoted

quoted

I see it is mostly busy with printing
WARNING: at .enable_kernel_fp+0x30/0x78, CPU#3: qemu-system-ppc/4884

The path seems to be same also.

[c000000085ca7700] [c00000000005de84] .kvmppc_load_guest_fp+0x30/0x80
[c000000085ca7780] [c00000000005f2a0] .kvmppc_handle_exit+0x5bc/0x5cc
[c000000085ca7830] [c00000000006204c] .kvmppc_resume_host+0xb8/0x10c


Does below help to boot or make it stable?
---
 arch/powerpc/kvm/booke.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
index 13ad4cf5fa71..d4be7a306bfc 100644
--- a/arch/powerpc/kvm/booke.c
+++ b/arch/powerpc/kvm/booke.c

@@ -1404,10 +1404,17 @@ int kvmppc_handle_exit(struct kvm_vcpu *vcpu, unsigned int exit_nr)

         if (s <= 0)
             r = (s << 2) | RESUME_HOST | (r & RESUME_FLAG_NV);
         else {
-            /* interrupts now hard-disabled */
+            /*
+             * interrupts now hard-disabled. i.e EE=0
+             * But irqs_disabled relies on PACA entries
+             * and fixup below enables it.
+             * So irqs_disabled == false
+             */
             kvmppc_fix_ee_before_entry();
+            preempt_disable();
             kvmppc_load_guest_fp(vcpu);
             kvmppc_load_guest_altivec(vcpu);
+            preempt_enable_no_resched();
         }
     }

Hi,

Thanks a lot for your patch!

I patched the 7.3-rc4 kernel sources with your new patch and compiled the kernel again yesterday. [1]

Thanks for trying.

quoted

Unfortunately, the issue still exists.

Can you share the console log of the kernel with the patch applied?
We need to see where it is stuck now. Has it made forward progress?

quoted

I tested it with a guest kernel with and without preempt support.

Cheers,
Christian

[1] https://github.com/chzigotzky/kernels/releases/tag/v7.3.0-rc4-2-preempt

Here is a new serial log from the patched kernel: https://github.com/user-attachments/files/32570808/e5500_preempt_pcie_ports_compat_patched_putty.log

That paca mismatch is independent of preemption modes.

It shows a few warnings and even after that a console message has come through.
Do warning continue forever?

To fix those warnings, can you try this diff on the top of the earlier ones?
(Igonore whitespace mangling if any, Also patch is UNTESTED. It is more of my
speculation.).

Sorry, I don't have any way to test it out locally.

---

diff --git a/arch/powerpc/kvm/powerpc.c b/arch/powerpc/kvm/powerpc.c
index 9194cf492d1c..0a014fa45b20 100644
--- a/arch/powerpc/kvm/powerpc.c
+++ b/arch/powerpc/kvm/powerpc.c

@@ -147,6 +147,19 @@ int kvmppc_prepare_to_enter(struct kvm_vcpu *vcpu)

                        continue;
                }

+#ifdef CONFIG_PPC64
+               if (unlikely(local_paca->irq_happened != PACA_IRQ_HARD_DIS)) {
+               /*
+                * An interrupt arrived between the soft and hard
+                * disable. Replay it and retry guest entry.
+                */
+                       local_irq_enable();
+                       local_irq_disable();
+                       hard_irq_disable();
+                       continue;
+               }
+#endif
+
                guest_enter_irqoff();
                return 1;
        }

Hi,

I patched the RC4 of kernel 7.3 with the two patches and compiled it again today.

I'm happy to report that the patched PREEMPT kernel with KVM HV works on the X5000/40. (e5500)

That's very nice to hear. I will write changelog for both and send it across soon.
Give me sometime.

Serial log: https://github.com/user-attachments/files/32619259/e5500_preempt_pcie_ports_compat_patched-2_putty.log

I also successfully tested it with KVM PR on the X1000 today. (PA6T)

Nice.

Further information: https://github.com/chzigotzky/kernels/issues/19#issuecomment-5818925408

Many thanks for your help,

Happy to help.
Thanks for reporting and testing. We can finally close this thread :)
Thanks to michal for helping in getting the logs.

Christian

Also, Michal, can you report that kvm specific issue you are seeing with the
ftrace fix applied in separate thread. This thread has become long already.