Page 104 of 124
Re: New kernels
Posted: Sun Jan 05, 2025 3:05 pm
by xeno74
Skateman wrote: Sun Jan 05, 2025 11:49 am
Kernel 6.13-RC5-2 up and running on my AmigaOne X5000
I have seen my desktop icons disapear and reappear a few times now..
Thank you for testing the RC5-2.
Maybe Caja doesn’t like our new kernels. it uses a lot of CPU resources (40%) after the first login.
After a logout and login, Caja works with low CPU resources again.
Please use top, htop or the MATE System Monitor to see if Caja is the problem.
Re: New kernels
Posted: Mon Jan 06, 2025 12:15 am
by kilaueabart
System Monitor says I am using 6.13.0-rc5. I downloaded uImage-6.13-rc5-kvm but I am getting too senile to know what to do with it, so no rc5-2. I did try popping it in alongside uImage-6.13, but I didn't expect it to change anything.
Re: New kernels
Posted: Mon Jan 06, 2025 4:30 am
by xeno74
kilaueabart wrote: Mon Jan 06, 2025 12:15 am
System Monitor says I am using 6.13.0-rc5. I downloaded uImage-6.13-rc5-kvm but I am getting too senile to know what to do with it, so no rc5-2. I did try popping it in alongside uImage-6.13, but I didn't expect it to change anything.
It is OK only to test the
RC5 if you don’t need KVM.
Thank you for testing!
Re: New kernels
Posted: Mon Jan 06, 2025 11:05 am
by xeno74
FYI:
NTFS mount without the NTFS3 kernel module:
df -Th
Code: Select all
/dev/sda1 fuseblk 1023 6.7M 1017M 1% /run/media/amigaone/NTFS_VOL
With NTFS3 kernel module:
Code: Select all
/dev/sda1 ntfs3 1023 6.7M 1017M 1% /run/media/amigaone/NTFS_VOL
Re: New kernels
Posted: Mon Jan 06, 2025 12:48 pm
by xeno74
Hi All,
Here is the
RC6 of kernel
6.13 for the X1000 and X5000.
Download and further information:
github.com
Please test the kernels.
Thanks,
Christian
Re: New kernels
Posted: Tue Jan 07, 2025 4:51 am
by xeno74
Alternative download link:
linux-image-6.13-rc6-X1000_X5000.tar.gz (xenosoft.de)
Re: New kernels
Posted: Wed Jan 08, 2025 10:47 am
by xeno74
FYI because of the KVM HV issue:
“Sean Christopherson“ wrote:
I would prefer to keep the sanity check to minimize the risk of a page fault handler not supporting opportunistic write mappings. e500 is definitely the odd one out here.
What about adding a dedicated wrapper for getting a writable PFN? E.g. (untested)
kvm_wrapper.patch:
Code: Select all
diff -rupN a/arch/powerpc/kvm/e500_mmu_host.c b/arch/powerpc/kvm/e500_mmu_host.c
--- a/arch/powerpc/kvm/e500_mmu_host.c 2025-01-05 23:13:40.000000000 +0100
+++ b/arch/powerpc/kvm/e500_mmu_host.c 2025-01-08 07:23:50.786895563 +0100
@@ -444,7 +444,7 @@ static inline int kvmppc_e500_shadow_map
if (likely(!pfnmap)) {
tsize_pages = 1UL << (tsize + 10 - PAGE_SHIFT);
- pfn = __kvm_faultin_pfn(slot, gfn, FOLL_WRITE, NULL, &page);
+ pfn = kvm_faultin_writable_pfn(slot, gfn, &page);
if (is_error_noslot_pfn(pfn)) {
if (printk_ratelimit())
pr_err("%s: real page not found for gfn %lx\n",
diff -rupN a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c
--- a/arch/x86/kvm/vmx/vmx.c 2025-01-05 23:13:40.000000000 +0100
+++ b/arch/x86/kvm/vmx/vmx.c 2025-01-08 07:28:21.392045189 +0100
@@ -6800,7 +6800,6 @@ void vmx_set_apic_access_page_addr(struc
struct page *refcounted_page;
unsigned long mmu_seq;
kvm_pfn_t pfn;
- bool writable;
/* Defer reload until vmcs01 is the current VMCS. */
if (is_guest_mode(vcpu)) {
@@ -6836,7 +6835,7 @@ void vmx_set_apic_access_page_addr(struc
* controls the APIC-access page memslot, and only deletes the memslot
* if APICv is permanently inhibited, i.e. the memslot won't reappear.
*/
- pfn = __kvm_faultin_pfn(slot, gfn, FOLL_WRITE, &writable, &refcounted_page);
+ pfn = kvm_faultin_writable_pfn(slot, gfn, &refcounted_page);
if (is_error_noslot_pfn(pfn))
return;
diff -rupN a/include/linux/kvm_host.h b/include/linux/kvm_host.h
--- a/include/linux/kvm_host.h 2025-01-05 23:13:40.000000000 +0100
+++ b/include/linux/kvm_host.h 2025-01-08 07:30:38.770596975 +0100
@@ -1276,6 +1276,14 @@ static inline kvm_pfn_t kvm_faultin_pfn(
write ? FOLL_WRITE : 0, writable, refcounted_page);
}
+static inline kvm_pfn_t kvm_faultin_writable_pfn(const struct kvm_memory_slot *slot,
+ gfn_t gfn, struct page **refcounted_page)
+{
+ bool writable;
+
+ return __kvm_faultin_pfn(slot, gfn, FOLL_WRITE, &writable, refcounted_page);
+}
+
int kvm_read_guest_page(struct kvm *kvm, gfn_t gfn, void *data, int offset,
int len);
int kvm_read_guest(struct kvm *kvm, gpa_t gpa, void *data, unsigned long len);
I have created a new RC6 of kernel 6.13 with this patch.
Download and further information:
github.com
Re: New kernels
Posted: Thu Jan 09, 2025 7:42 pm
by xeno74
Paolo has released a patch series for solving the KVM HV issue.
I compiled the RC6 of kernel 6.13 with Paolo‘s patch series today.
Download and further information:
github.com
Re: New kernels
Posted: Sat Jan 11, 2025 12:49 pm
by xeno74
I tested the
RC6 of kernel
6.13 on my X5000/40 today. KVM HV works without any problems with the
kvm_hv_final.patch.
Re: New kernels
Posted: Mon Jan 13, 2025 2:27 pm
by xeno74
Hi All,
I released the
RC7 of kernel
6.13 for the X1000 and X5000 today.
Download and further information:
github.com
Please test the kernels.
Thanks,
Christian
