On 6/13/24 1:04 PM, Jan Beulich wrote:
On 24.05.2024 14:31, Vaishali Thakkar wrote:
--- a/xen/arch/x86/hvm/asid.c
+++ b/xen/arch/x86/hvm/asid.c
@@ -27,8 +27,8 @@ boolean_param("asid", opt_asid_enabled);
* the TLB.
*
* Sketch of the Implementation:
- *
- * ASIDs are a CPU-loca
ested on AMD EPYC Genoa at the moment.
Vaishali Thakkar (1):
x86/hvm: Introduce Xen-wide ASID allocator
xen/arch/x86/domain.c | 7 ++
xen/arch/x86/flushtlb.c| 4 -
xen/arch/x86/hvm/asid.c| 136 +
xen/arch/x86/hvm/hvm.c
rt of xen and allows to have a fixed asids
throughout the lifecycle of all domains. Having a fixed asid
for non-SEV domains also presents us with the opportunity to
further take use of AMD instructions like TLBSYNC and INVLPGB
for broadcasting the TLB invalidations.
Signed-off-by: Vaishali Th
rt of xen and allows to have a fixed asids
throughout the lifecycle of all domains. Having a fixed asid
for non-SEV domains also presents us with the opportunity to
further take use of AMD instructions like TLBSYNC and INVLPGB
for broadcasting the TLB invalidations.
Signed-off-by: Vaishali Th
o pick the asid from such stack before doing
hvm_asid_flush_all.
[0]
https://www.amd.com/content/dam/amd/en/documents/processor-tech-docs/programmer-references/24593.pdf
Vaishali Thakkar (1):
x86/hvm: Introduce Xen-wide ASID Allocator
xen/arch/x86/flushtlb.c| 1 -
xen/ar
Hi All,
This is an informational post about our plans for the enablement of AMD SEV
support in Xen. This work will be done as part of the Hyper OpenX project[1].
Phase Zero:
Our primary intention is to gather the necessary information required to
commence the upstream work f
The suffix is redundant for np/sev/sev-es bits. Drop it
to avoid adding extra code volume.
Suggested-by: Andrew Cooper
Signed-off-by: Vaishali Thakkar i
---
xen/arch/x86/hvm/svm/nestedsvm.c| 14 +++---
xen/arch/x86/hvm/svm/svm.c | 2 +-
xen/arch/x86/hvm/svm/vmcb.c
While sev and sev_es bits are not yet enabled in xen,
including their status in the VMCB dump could be
informational.Therefore, print it via svmdebug.
Signed-off-by: Vaishali Thakkar
---
JFYI, we'll send the follow-up patches with the enablement
of sev and ASP driver.
---
xen/arch/x86/hv
Hi,
In this patchset, first patch removes the unnecessary suffix
from a bunch of vmcb bits and the second patch is about
printing the status of sev and sev-es bits while dumping VMCB.
Vaishali Thakkar (2):
x86/svm: Drop the _enabled suffix from vmcb bits
x86/svmdebug: Print sev and sev_es
On 3/8/24 00:22, Andrew Cooper wrote:
On 07/03/2024 9:40 pm, Vaishali Thakkar wrote:
The suffix is redundant for np/sev/sev-es bits. Drop it
to avoid adding extra code volume.
Suggested-by: Andrew Cooper
Signed-off-by: Vaishali Thakkar i
Typo on the end of your email address?
Oops, thanks
On 3/8/24 00:34, Andrew Cooper wrote:
On 07/03/2024 9:40 pm, Vaishali Thakkar wrote:
diff --git a/xen/arch/x86/hvm/svm/svmdebug.c b/xen/arch/x86/hvm/svm/svmdebug.c
index 24358c6eea..f54b426fb3 100644
--- a/xen/arch/x86/hvm/svm/svmdebug.c
+++ b/xen/arch/x86/hvm/svm/svmdebug.c
@@ -53,6 +53,8
per Andrew's
suggestion in one of the reviews
- Address Andrew's comment with respect to pretty printing
Vaishali Thakkar (3):
x86/svm: Drop the _enabled suffix from vmcb bits
x86/svm: Drop the suffix _guest from vmcb bit
x86/svmdebug: Print sev and sev_es vmcb bits
xen/ar
The suffix _guest is redundant for asid bit. Drop it
to avoid adding extra code volume.
Suggested-by: Andrew Cooper
Signed-off-by: Vaishali Thakkar
---
Changes since v1:
- This patch wasn't part of v1. It's been added to
address Andrew's suggestion.
---
xen/a
While sev and sev_es bits are not yet enabled in xen,
including their status in the VMCB dump could be
informational.Therefore, print it via svmdebug.
Signed-off-by: Vaishali Thakkar
---
Changes since v1:
- Pretty printing
---
xen/arch/x86/hvm/svm/svmdebug.c | 7 +--
1 file changed
The suffix is redundant for np/sev/sev-es bits. Drop it
to avoid adding extra code volume. While we're here, drop
the double negations in one of the instances of _np bit
and replace 0/1 with false/true in the use cases of _np.
Suggested-by: Andrew Cooper
Signed-off-by: Vaishali Thakkar
Rev
On 3/12/24 08:54, Jan Beulich wrote:
On 11.03.2024 13:40, Vaishali Thakkar wrote:
--- a/xen/arch/x86/hvm/svm/nestedsvm.c
+++ b/xen/arch/x86/hvm/svm/nestedsvm.c
@@ -571,7 +571,7 @@ static int nsvm_vmcb_prepare4vmrun(struct vcpu *v, struct
cpu_user_regs *regs)
if
On 3/12/24 08:59, Jan Beulich wrote:
On 11.03.2024 13:40, Vaishali Thakkar wrote:
@@ -698,11 +698,11 @@ nsvm_vcpu_vmentry(struct vcpu *v, struct cpu_user_regs
*regs,
/* Convert explicitely to boolean. Deals with l1 guests
* that use flush-by-asid w/o checking the cpuid bits
On 3/12/24 11:49, Jan Beulich wrote:
On 12.03.2024 11:00, Vaishali Thakkar wrote:
On 3/12/24 08:54, Jan Beulich wrote:
On 11.03.2024 13:40, Vaishali Thakkar wrote:
--- a/xen/arch/x86/hvm/svm/nestedsvm.c
+++ b/xen/arch/x86/hvm/svm/nestedsvm.c
@@ -571,7 +571,7 @@ static int
On 3/12/24 09:05, Jan Beulich wrote:
On 11.03.2024 13:40, Vaishali Thakkar wrote:
While sev and sev_es bits are not yet enabled in xen,
including their status in the VMCB dump could be
informational.Therefore, print it via svmdebug.
Yet there are more bits there. I'm okay with leavin
he pretty printing in svm-debug based on Jan's
comments
- Improve commit logs and the subject of patch 3 to include
the changes done in v3
Vaishali Thakkar (3):
x86/svm: Drop the _enabled suffix from vmcb bits
x86/svm: Drop the suffix _guest from vmcb bit
x86/svmdebug: Print
The suffix _guest is redundant for asid bit. Drop it
to avoid adding extra code volume.
While we're here, replace 0/1 with false/true and use
VMCB accessors instead of open coding.
Suggested-by: Andrew Cooper
Signed-off-by: Vaishali Thakkar
---
Changes since v1:
- This patch w
ndrew Cooper
Signed-off-by: Vaishali Thakkar
Reviewed-by: Andrew Cooper
---
Changes since v1:
- Address Andrew and Jan's reviews related to dropping
double negation and replacing 0/1 with false/true
- Fix the typo around signed-off-by
Changes since v2:
is in
progress. And it'll be useful to have this information as
part of svmdebug.
Signed-off-by: Vaishali Thakkar
---
Changes since v1:
- Pretty printing
Changes since v2:
- Minor changes in pretty printing to make information
clear
- Improve commit log and su
On 3/14/24 17:04, Jan Beulich wrote:
On 13.03.2024 17:41, Vaishali Thakkar wrote:
The suffix _guest is redundant for asid bit. Drop it
to avoid adding extra code volume.
While we're here, replace 0/1 with false/true and use
VMCB accessors instead of open coding.
Suggested-by: Andrew C
On 4/12/24 4:06 PM, Andrei Semenov wrote:
On 4/11/24 20:32, Andrew Cooper wrote:
On 10/04/2024 4:36 pm, Andrei Semenov wrote:
diff --git a/xen/arch/x86/cpu/amd.c b/xen/arch/x86/cpu/amd.c
index ab92333673..a5903613f0 100644
--- a/xen/arch/x86/cpu/amd.c
+++ b/xen/arch/x86/cpu/amd.c
@@ -1030,6 +1
On 4/12/24 5:07 PM, Andrew Cooper wrote:
On 12/04/2024 3:38 pm, Vaishali Thakkar wrote:
On 4/12/24 4:06 PM, Andrei Semenov wrote:
On 4/11/24 20:32, Andrew Cooper wrote:
On 10/04/2024 4:36 pm, Andrei Semenov wrote:
+ }
+
+ if (!(cpu_has_sme || cpu_has_sev))
+ return;
+
+ if
, add min_asid
to allow supplying the dynamic start ASID during the
allocation process.
Signed-off-by: Vaishali Thakkar
---
xen/arch/x86/hvm/asid.c | 10 ++
1 file changed, 6 insertions(+), 4 deletions(-)
diff --git a/xen/arch/x86/hvm/asid.c b/xen/arch/x86/hvm/asid.c
index 8d27b7dba1..e14b6
TLB Flush by ASID is missing in the list of supported features
here. So, add it.
Signed-off-by: Vaishali Thakkar
---
xen/arch/x86/hvm/svm/svm.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/xen/arch/x86/hvm/svm/svm.c b/xen/arch/x86/hvm/svm/svm.c
index a745acd903..4719fffae5 100644
--- a
On 4/16/24 3:38 PM, Andrew Cooper wrote:
On 16/04/2024 10:08 am, Vaishali Thakkar wrote:
TLB Flush by ASID is missing in the list of supported features
here. So, add it.
Signed-off-by: Vaishali Thakkar
---
xen/arch/x86/hvm/svm/svm.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/xen
On 4/16/24 4:12 PM, Andrew Cooper wrote:
On 16/04/2024 9:54 am, Vaishali Thakkar wrote:
Currently, Xen always starts the ASID allocation at 1. But
for SEV technologies the ASID space is divided. This is
because it's a security issue if a guest is started as
ES/SNP and is migrated to SEV
On 4/16/24 4:25 PM, Vaishali Thakkar wrote:
On 4/16/24 4:12 PM, Andrew Cooper wrote:
On 16/04/2024 9:54 am, Vaishali Thakkar wrote:
Currently, Xen always starts the ASID allocation at 1. But
for SEV technologies the ASID space is divided. This is
because it's a security issue if a gue
ain.c which need the header. Clean up all others.
No functional change.
Signed-off-by: Andrew Cooper
Reviewed-by: Vaishali Thakkar
---
CC: Jan Beulich
CC: Roger Pau Monné
CC: Stefano Stabellini
CC: Xenia Ragiadakou
CC: Sergiy Kibrik
CC: George Dunlap
CC: Andrei Semenov
CC: Vaishali Th
On 4/29/24 5:16 PM, Andrew Cooper wrote:
For display purposes only right now.
Signed-off-by: Andrew Cooper
Reviewed-by: Vaishali Thakkar
---
CC: Jan Beulich
CC: Roger Pau Monné
CC: Stefano Stabellini
CC: Xenia Ragiadakou
CC: Sergiy Kibrik
CC: George Dunlap
CC: Andrei Semenov
CC
Hi All,
Gentle ping, I would appreciate some reviews here. Especially because in v2
the whole domain liefcylce with the fixed ASID works. So, I would love to
know if it's going in the right direction or not.
Thanks!
On 8/13/24 8:58 PM, Vaishali Thakkar wrote:
Motivation:
---
Th
34 matches
Mail list logo
