pool/qemu
SHA256
6
0
Fork 1
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Accepting request 676606 from home:bfrogers:branches:Virtualization

Browse Source 
- Patch queue updated from git://github.com/openSUSE/qemu.git opensuse-3.1
* Patches added:
  0058-Revert-target-i386-kvm-add-VMX-migr.patch
  0059-memory-Fix-the-memory-region-type-a.patch
  0060-target-i386-sev-Do-not-pin-the-ram-.patch
- Revert upstream patch which declares x86 vmx feature a migration
  blocker. Given the proliferation of using vm's with host features
  passed through and the general knowledge that nested
  virtualization has many usage caveats, but still gets put in use
  in restricted scenarios, this patch did more harm than good, I
  feel. So despite this relaxation, please consider yourself warned
  that nested virtualization is not yet a supportable feature.
  (bsc#1121604) 
  0058-Revert-target-i386-kvm-add-VMX-migr.patch
- Fix SEV VM device assignment (bsc#1123205)
  0059-memory-Fix-the-memory-region-type-a.patch
  0060-target-i386-sev-Do-not-pin-the-ram-.patch
- Patch queue updated from git://github.com/openSUSE/qemu.git opensuse-3.1
- Revert upstream patch which declares x86 vmx feature a migration
  blocker. Given the proliferation of using vm's with host features
  passed through and the general knowledge that nested
  virtualization has many usage caveats, but still gets put in use
  in restricted scenarios, this patch did more harm than good, I
  feel. So despite this relaxation, please consider yourself warned
  that nested virtualization is not yet a supportable feature.
  (bsc#1121604) 
  0058-Revert-target-i386-kvm-add-VMX-migr.patch
- Fix SEV VM device assignment (bsc#1123205)
  0059-memory-Fix-the-memory-region-type-a.patch
  0060-target-i386-sev-Do-not-pin-the-ram-.patch
- Patch queue updated from git://github.com/openSUSE/qemu.git opensuse-3.1

OBS-URL: https://build.opensuse.org/request/show/676606
OBS-URL: https://build.opensuse.org/package/show/Virtualization/qemu?expand=0&rev=455
This commit is contained in:
Bruce Rogers 2019-02-15 22:57:45 +00:00 committed by Git OBS Bridge
parent 367159087c
commit a9eb5a2a28
9 changed files with 213 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

65
0058-Revert-target-i386-kvm-add-VMX-migr.patch Normal file
View File

@ -0,0 +1,65 @@
From: Bruce Rogers <brogers@suse.com>
Date: Fri, 15 Feb 2019 15:12:04 -0700
Subject: Revert "target/i386: kvm: add VMX migration blocker"
This reverts commit d98f26073bebddcd3da0ba1b86c3a34e840c0fb8.
Here is some text explaining the revert:
I've thought about this some more, and with upstream
discussions about it having stagnated, at this point I think
the best solution is to revert the patch which considers it
a migration blocker to have the vmx feature enabled. It's
worth noting that not only are migrations blocked, but
saving of the vm state via save/restore and snapshots.
Given that it is still widely known that Nested Virtualization
is not supported by SUSE and other vendors, but is still used
by quite a few people who understand that there are caveats
with it's usage, I believe this migration blocker is more
hurtful than helpful.
The fact that as of the v4.20 kernel, nested virtualization is
enabled by default (for vmx), was partly why the patch was
added in the first place. But my perspective is that perhaps
enabling nested was still a bit premature.
I will make sure our qemu changelog explains that despite
removing that migration blocker, the user is warned that
nested virtualization is still a "use at your own risk
feature".
[BR: BSC#1121604]
Signed-off-by: Bruce Rogers <brogers@suse.com>
---
target/i386/kvm.c | 12 ------------
1 file changed, 12 deletions(-)
diff --git a/target/i386/kvm.c b/target/i386/kvm.c
index b2401d13ea..f97bfc164d 100644
--- a/target/i386/kvm.c
+++ b/target/i386/kvm.c
@@ -855,7 +855,6 @@ static int hyperv_init_vcpu(X86CPU *cpu)
}
static Error *invtsc_mig_blocker;
-static Error *vmx_mig_blocker;
#define KVM_MAX_CPUID_ENTRIES 100
@@ -1248,17 +1247,6 @@ int kvm_arch_init_vcpu(CPUState *cs)
!!(c->ecx & CPUID_EXT_SMX);
}
- if ((env->features[FEAT_1_ECX] & CPUID_EXT_VMX) && !vmx_mig_blocker) {
- error_setg(&vmx_mig_blocker,
- "Nested VMX virtualization does not support live migration yet");
- r = migrate_add_blocker(vmx_mig_blocker, &local_err);
- if (local_err) {
- error_report_err(local_err);
- error_free(vmx_mig_blocker);
- return r;
- }
- }
-
if (env->mcg_cap & MCG_LMCE_P) {
has_msr_mcg_ext_ctl = has_msr_feature_control = true;
}

44
0059-memory-Fix-the-memory-region-type-a.patch Normal file
View File

@ -0,0 +1,44 @@
From: "Singh, Brijesh" <brijesh.singh@amd.com>
Date: Fri, 15 Feb 2019 14:22:21 -0700
Subject: memory: Fix the memory region type assignment order
Currently, a callback registered through the RAMBlock notifier
is not able to get the memory region type (i.e callback is not
able to use memory_region_is_ram_device function). This is
because mr->ram assignment happens _after_ the memory is allocated
whereas the callback is executed during allocation.
Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1667249
Suggested-by: Alex Williamson <alex.williamson@redhat.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Brijesh Singh <brijesh.singh@amd.com>
[BSC#1123205]
Signed-off-by: Bruce Rogers <brogers@suse.com>
---
memory.c | 9 ++++++++-
1 file changed, 8 insertions(+), 1 deletion(-)
diff --git a/memory.c b/memory.c
index d14c6dec1d..d1b68fdae8 100644
--- a/memory.c
+++ b/memory.c
@@ -1605,10 +1605,17 @@ void memory_region_init_ram_device_ptr(MemoryRegion *mr,
uint64_t size,
void *ptr)
{
- memory_region_init_ram_ptr(mr, owner, name, size, ptr);
+ memory_region_init(mr, owner, name, size);
+ mr->ram = true;
+ mr->terminates = true;
mr->ram_device = true;
mr->ops = &ram_device_mem_ops;
mr->opaque = mr;
+ mr->destructor = memory_region_destructor_ram;
+ mr->dirty_log_mask = tcg_enabled() ? (1 << DIRTY_MEMORY_CODE) : 0;
+ /* qemu_ram_alloc_from_ptr cannot fail with ptr != NULL. */
+ assert(ptr != NULL);
+ mr->ram_block = qemu_ram_alloc_from_ptr(size, ptr, mr, &error_fatal);
}
void memory_region_init_alias(MemoryRegion *mr,

43
0060-target-i386-sev-Do-not-pin-the-ram-.patch Normal file
View File

@ -0,0 +1,43 @@
From: "Singh, Brijesh" <brijesh.singh@amd.com>
Date: Fri, 15 Feb 2019 14:24:21 -0700
Subject: target/i386: sev: Do not pin the ram device memory region
The RAM device presents a memory region that should be handled
as an IO region and should not be pinned.
In the case of the vfio-pci, RAM device represents a MMIO BAR
and the memory region is not backed by pages hence
KVM_MEMORY_ENCRYPT_REG_REGION fails to lock the memory range.
Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1667249
Cc: Alex Williamson <alex.williamson@redhat.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Brijesh Singh <brijesh.singh@amd.com>
[BSC#1123205]
Signed-off-by: Bruce Rogers <brogers@suse.com>
---
target/i386/sev.c | 11 +++++++++++
1 file changed, 11 insertions(+)
diff --git a/target/i386/sev.c b/target/i386/sev.c
index 2395171acf..7d6f4032d6 100644
--- a/target/i386/sev.c
+++ b/target/i386/sev.c
@@ -130,6 +130,17 @@ sev_ram_block_added(RAMBlockNotifier *n, void *host, size_t size)
{
int r;
struct kvm_enc_region range;
+ ram_addr_t offset;
+ MemoryRegion *mr;
+
+ /*
+ * The RAM device presents a memory region that should be treated
+ * as IO region and should not be pinned.
+ */
+ mr = memory_region_from_host(host, &offset);
+ if (mr && memory_region_is_ram_device(mr)) {
+ return;
+ }
range.addr = (__u64)(unsigned long)host;
range.size = size;

9
qemu-linux-user.changes
View File

@ -1,3 +1,12 @@
-------------------------------------------------------------------
Fri Feb 15 22:49:26 UTC 2019 - Bruce Rogers <brogers@suse.com>
- Patch queue updated from git://github.com/openSUSE/qemu.git opensuse-3.1
* Patches added:
0058-Revert-target-i386-kvm-add-VMX-migr.patch
0059-memory-Fix-the-memory-region-type-a.patch
0060-target-i386-sev-Do-not-pin-the-ram-.patch
-------------------------------------------------------------------
Wed Jan 30 15:54:31 UTC 2019 - Liang Yan <lyan@suse.com>

6
qemu-linux-user.spec
View File

@ -89,6 +89,9 @@ Patch0054: 0054-linux-user-make-pwrite64-pread64-fd.patch
Patch0055: 0055-xen-Add-xen-v4.12-based-xc_domain_c.patch
Patch0056: 0056-slirp-check-data-length-while-emula.patch
Patch0057: 0057-s390x-Return-specification-exceptio.patch
Patch0058: 0058-Revert-target-i386-kvm-add-VMX-migr.patch
Patch0059: 0059-memory-Fix-the-memory-region-type-a.patch
Patch0060: 0060-target-i386-sev-Do-not-pin-the-ram-.patch
# Please do not add QEMU patches manually here.
# Run update_git.sh to regenerate this queue.
ExcludeArch: s390
@ -177,6 +180,9 @@ syscall layer occurs on the native hardware and operating system.
%patch0055 -p1
%patch0056 -p1
%patch0057 -p1
%patch0058 -p1
%patch0059 -p1
%patch0060 -p1
%build
./configure \

17
qemu-testsuite.changes
View File

@ -1,3 +1,20 @@
-------------------------------------------------------------------
Fri Feb 15 22:49:24 UTC 2019 - Bruce Rogers <brogers@suse.com>
- Revert upstream patch which declares x86 vmx feature a migration
blocker. Given the proliferation of using vm's with host features
passed through and the general knowledge that nested
virtualization has many usage caveats, but still gets put in use
in restricted scenarios, this patch did more harm than good, I
feel. So despite this relaxation, please consider yourself warned
that nested virtualization is not yet a supportable feature.
(bsc#1121604)
0058-Revert-target-i386-kvm-add-VMX-migr.patch
- Fix SEV VM device assignment (bsc#1123205)
0059-memory-Fix-the-memory-region-type-a.patch
0060-target-i386-sev-Do-not-pin-the-ram-.patch
- Patch queue updated from git://github.com/openSUSE/qemu.git opensuse-3.1
-------------------------------------------------------------------
Mon Feb 11 15:41:02 UTC 2019 - Bruce Rogers <brogers@suse.com>

6
qemu-testsuite.spec
View File

@ -193,6 +193,9 @@ Patch0054: 0054-linux-user-make-pwrite64-pread64-fd.patch
Patch0055: 0055-xen-Add-xen-v4.12-based-xc_domain_c.patch
Patch0056: 0056-slirp-check-data-length-while-emula.patch
Patch0057: 0057-s390x-Return-specification-exceptio.patch
Patch0058: 0058-Revert-target-i386-kvm-add-VMX-migr.patch
Patch0059: 0059-memory-Fix-the-memory-region-type-a.patch
Patch0060: 0060-target-i386-sev-Do-not-pin-the-ram-.patch
# Please do not add QEMU patches manually here.
# Run update_git.sh to regenerate this queue.
@ -980,6 +983,9 @@ This package provides a service file for starting and stopping KSM.
%patch0055 -p1
%patch0056 -p1
%patch0057 -p1
%patch0058 -p1
%patch0059 -p1
%patch0060 -p1
pushd roms/seabios
%patch1100 -p1

17
qemu.changes
View File

@ -1,3 +1,20 @@
-------------------------------------------------------------------
Fri Feb 15 22:49:24 UTC 2019 - Bruce Rogers <brogers@suse.com>
- Revert upstream patch which declares x86 vmx feature a migration
blocker. Given the proliferation of using vm's with host features
passed through and the general knowledge that nested
virtualization has many usage caveats, but still gets put in use
in restricted scenarios, this patch did more harm than good, I
feel. So despite this relaxation, please consider yourself warned
that nested virtualization is not yet a supportable feature.
(bsc#1121604)
0058-Revert-target-i386-kvm-add-VMX-migr.patch
- Fix SEV VM device assignment (bsc#1123205)
0059-memory-Fix-the-memory-region-type-a.patch
0060-target-i386-sev-Do-not-pin-the-ram-.patch
- Patch queue updated from git://github.com/openSUSE/qemu.git opensuse-3.1
-------------------------------------------------------------------
Mon Feb 11 15:41:02 UTC 2019 - Bruce Rogers <brogers@suse.com>

6
qemu.spec
View File

@ -193,6 +193,9 @@ Patch0054: 0054-linux-user-make-pwrite64-pread64-fd.patch
Patch0055: 0055-xen-Add-xen-v4.12-based-xc_domain_c.patch
Patch0056: 0056-slirp-check-data-length-while-emula.patch
Patch0057: 0057-s390x-Return-specification-exceptio.patch
Patch0058: 0058-Revert-target-i386-kvm-add-VMX-migr.patch
Patch0059: 0059-memory-Fix-the-memory-region-type-a.patch
Patch0060: 0060-target-i386-sev-Do-not-pin-the-ram-.patch
# Please do not add QEMU patches manually here.
# Run update_git.sh to regenerate this queue.
@ -980,6 +983,9 @@ This package provides a service file for starting and stopping KSM.
%patch0055 -p1
%patch0056 -p1
%patch0057 -p1
%patch0058 -p1
%patch0059 -p1
%patch0060 -p1
pushd roms/seabios
%patch1100 -p1