On Thu, Aug 31, 2017 at 02:33:57AM -0600, Jan Beulich wrote:
On 31.08.17 at 09:15, wrote:
>> On Thu, Aug 31, 2017 at 01:42:53AM -0600, Jan Beulich wrote:
>> On 31.08.17 at 00:57, wrote:
On Wed, Aug 30, 2017 at 10:00:49AM -0600, Jan Beulich wrote:
On 16.08.17 at 07:14, wrot
On Thu, Aug 31, 2017 at 10:03:19AM +0100, Roger Pau Monne wrote:
>On Thu, Aug 31, 2017 at 03:32:42PM +0800, Chao Gao wrote:
>> On Tue, Aug 29, 2017 at 08:33:25AM +0100, Roger Pau Monne wrote:
>> >On Mon, Aug 28, 2017 at 06:18:13AM +, Tian, Kevin wrote:
>> &g
On Thu, Aug 31, 2017 at 02:33:57AM -0600, Jan Beulich wrote:
On 31.08.17 at 09:15, wrote:
>> On Thu, Aug 31, 2017 at 01:42:53AM -0600, Jan Beulich wrote:
>> On 31.08.17 at 00:57, wrote:
On Wed, Aug 30, 2017 at 10:00:49AM -0600, Jan Beulich wrote:
On 16.08.17 at 07:14, wrot
On Fri, Sep 01, 2017 at 02:24:08AM -0600, Jan Beulich wrote:
On 01.09.17 at 03:39, wrote:
>> After thinking it again, I want to define the counter as
>> a unsigned int variable for the following reasion:
>> 1. It is definite that the counter is closely related with
>> list_add() and list_del(
On Fri, Sep 01, 2017 at 03:13:17AM -0600, Jan Beulich wrote:
On 01.09.17 at 09:55, wrote:
>> On Fri, Sep 01, 2017 at 02:24:08AM -0600, Jan Beulich wrote:
>> On 01.09.17 at 03:39, wrote:
After thinking it again, I want to define the counter as
a unsigned int variable for the fol
On Thu, Aug 31, 2017 at 11:09:48AM +0100, Roger Pau Monne wrote:
>On Thu, Aug 31, 2017 at 04:45:23PM +0800, Chao Gao wrote:
>> On Thu, Aug 31, 2017 at 10:03:19AM +0100, Roger Pau Monne wrote:
>> >On Thu, Aug 31, 2017 at 03:32:42PM +0800, Chao Gao wrote:
>> >> On T
On Mon, Sep 04, 2017 at 10:26:04AM +0100, Roger Pau Monné wrote:
>(Adding Chao again because my MUA seems to drop him each time)
>
>On Mon, Sep 04, 2017 at 10:00:00AM +0100, Roger Pau Monné wrote:
>> On Mon, Sep 04, 2017 at 02:25:10PM +0800, Chao Gao wrote:
>> > On Thu, Au
Hi, Roger.
I meet an error when creating a pvh guest. I am using commit 6e2a4c73564a.
From the error log, I found bootlate_pv()->pin_table always failed. And the
failure was caused by is_pv_domain(pg_owner) in do_mmuext_op(). Do you have
any idea on this?
The guest config file is:
builder = "hvm
On Tue, Oct 03, 2017 at 11:08:01AM +0100, Roger Pau Monné wrote:
>On Tue, Oct 03, 2017 at 09:55:44AM +, osstest service owner wrote:
>> flight 113959 xen-unstable real [real]
>> http://logs.test-lab.xenproject.org/osstest/logs/113959/
>>
>> Regressions :-(
>>
>> Tests which did not succeed an
On Mon, Oct 09, 2017 at 02:13:22PM +0800, Chao Gao wrote:
>On Tue, Oct 03, 2017 at 11:08:01AM +0100, Roger Pau Monné wrote:
>>On Tue, Oct 03, 2017 at 09:55:44AM +, osstest service owner wrote:
>>> flight 113959 xen-unstable real [real]
>>> http://logs.test-lab.
On Mon, Oct 09, 2017 at 12:03:53PM +0100, Andrew Cooper wrote:
>On 09/10/17 08:58, Chao Gao wrote:
>> On Mon, Oct 09, 2017 at 02:13:22PM +0800, Chao Gao wrote:
>>> On Tue, Oct 03, 2017 at 11:08:01AM +0100, Roger Pau Monné wrote:
>>>> On Tue, Oct 03, 2017 at 09:55:44A
patch
holds the irq_lock when doing the two operations to prevent the case.
Signed-off-by: Chao Gao
---
xen/arch/x86/hvm/irq.c | 11 ++
xen/arch/x86/hvm/vlapic.c| 11 ++
xen/arch/x86/hvm/vpt.c | 43
xen/include/as
someone changes the vector field of IOAPIC RTE between asserting
>> the irq and getting the vector of the irq, leading to setting the
>> old vector number but returning a different vector number. This patch
>> holds the irq_lock when doing the two operations to prevent the case.
&
or number. This patch
allows hvm_isa_irq_assert() to accept a callback which can get the
interrupt vector with irq_lock held. Thus, no one can change the vector
between the two operations.
Signed-off-by: Chao Gao
---
passed the two simple xtf tests in
https://lists.xenproject.org/archives/html/xe
On Mon, Oct 16, 2017 at 07:15:16AM -0600, Jan Beulich wrote:
On 13.10.17 at 07:10, wrote:
>> --- a/xen/arch/x86/hvm/irq.c
>> +++ b/xen/arch/x86/hvm/irq.c
>> @@ -168,11 +168,13 @@ void hvm_gsi_deassert(struct domain *d, unsigned int
>> gsi)
>> spin_unlock(&d->arch.hvm_domain.irq_lock);
>
On Mon, Oct 16, 2017 at 08:26:09AM -0600, Jan Beulich wrote:
On 16.10.17 at 15:13, wrote:
>> On Mon, Oct 16, 2017 at 07:15:16AM -0600, Jan Beulich wrote:
>> On 13.10.17 at 07:10, wrote:
--- a/xen/arch/x86/hvm/irq.c
+++ b/xen/arch/x86/hvm/irq.c
@@ -168,11 +168,13 @@ void hv
On Thu, Oct 19, 2017 at 09:56:34AM -0600, Jan Beulich wrote:
On 19.10.17 at 17:49, wrote:
>> On Thu, Sep 21, 2017 at 11:02:03PM -0400, Lan Tianyu wrote:
>>> --- a/xen/arch/x86/hvm/vioapic.c
>>> +++ b/xen/arch/x86/hvm/vioapic.c
>>> @@ -561,11 +561,25 @@ int vioapic_get_vector(const struct doma
On Thu, Oct 19, 2017 at 10:49:22AM +0100, Roger Pau Monné wrote:
>On Thu, Sep 21, 2017 at 11:01:47PM -0400, Lan Tianyu wrote:
>> From: Chao Gao
>>
>> A field, viommu_info, is added to struct libxl_domain_build_info. Several
>> attributes can be specified by guest co
On Thu, Oct 19, 2017 at 11:00:27AM +0100, Roger Pau Monné wrote:
>On Thu, Sep 21, 2017 at 11:01:48PM -0400, Lan Tianyu wrote:
>> From: Chao Gao
>>
>> A new logic is added to build ACPI DMAR table in tool stack for a guest
>> with one virtual VTD and pass throu
On Thu, Oct 19, 2017 at 11:21:35AM +0100, Roger Pau Monné wrote:
>On Thu, Sep 21, 2017 at 11:01:51PM -0400, Lan Tianyu wrote:
>> From: Chao Gao
>>
>> No functional changes.
>>
>> Signed-off-by: Chao Gao
>> Signed-off-by: Lan Tianyu
>
>Reviewed-by:
On Thu, Oct 19, 2017 at 12:20:35PM +0100, Roger Pau Monné wrote:
>On Thu, Sep 21, 2017 at 11:01:52PM -0400, Lan Tianyu wrote:
>> From: Chao Gao
>>
>> This patch adds create/destroy function for the emulated VTD
>> and adapts it to the common VIOMMU abstraction.
>
On Thu, Oct 19, 2017 at 12:34:54PM +0100, Roger Pau Monné wrote:
>On Thu, Sep 21, 2017 at 11:01:53PM -0400, Lan Tianyu wrote:
>> From: Chao Gao
>>
>> This patch adds VVTD MMIO handler to deal with MMIO access.
>>
>> Signed-off-by: Chao Gao
>> Signed-off
On Thu, Oct 19, 2017 at 12:56:45PM +0100, Roger Pau Monné wrote:
>On Thu, Sep 21, 2017 at 11:01:54PM -0400, Lan Tianyu wrote:
>> From: Chao Gao
>>
>> Software sets this field to set/update the interrupt remapping table pointer
>> used by hardware. The interrupt
On Thu, Oct 19, 2017 at 03:26:30PM +0100, Roger Pau Monné wrote:
>On Thu, Sep 21, 2017 at 11:01:56PM -0400, Lan Tianyu wrote:
>> From: Chao Gao
>>
>> When a remapping interrupt request arrives, remapping hardware computes the
>> interrupt_index per the algo
On Thu, Oct 19, 2017 at 03:39:44PM +0100, Roger Pau Monné wrote:
>On Thu, Sep 21, 2017 at 11:01:57PM -0400, Lan Tianyu wrote:
>> From: Chao Gao
>>
>> Without interrupt remapping, interrupt attributes can be extracted from
>> msi message or IOAPIC RTE. However, with
On Thu, Oct 19, 2017 at 05:03:26PM +0100, Roger Pau Monné wrote:
>On Thu, Sep 21, 2017 at 11:02:05PM -0400, Lan Tianyu wrote:
>> From: Chao Gao
>
>The title for this patch it's wrong, it modifies both the hypervisor
>and libxc. Please fix it.
>
>> When exposing v
On Thu, Oct 19, 2017 at 05:31:37PM +0100, Roger Pau Monné wrote:
>On Thu, Sep 21, 2017 at 11:02:07PM -0400, Lan Tianyu wrote:
>> From: Chao Gao
>>
>> Interrupt translation faults are non-recoverable fault. When faults
>> are triggered, it needs to populate fa
On Fri, Oct 20, 2017 at 12:56:03AM -0600, Jan Beulich wrote:
>>>> On 20.10.17 at 04:46, wrote:
>> On Thu, Oct 19, 2017 at 12:20:35PM +0100, Roger Pau Monné wrote:
>>>On Thu, Sep 21, 2017 at 11:01:52PM -0400, Lan Tianyu wrote:
>>>> From: Chao Gao
>>>
or number. This patch
allows hvm_isa_irq_assert() to accept a callback which can get the
interrupt vector with irq_lock held. Thus, no one can change the vector
between the two operations.
BTW, the first argument of pi_test_and_set_pir() should be uint8_t
and I take this chance to fix it.
Signed-o
On Fri, Oct 20, 2017 at 11:01:03AM +0100, Roger Pau Monné wrote:
>On Fri, Oct 20, 2017 at 01:16:37PM +0800, Chao Gao wrote:
>> On Thu, Oct 19, 2017 at 03:26:30PM +0100, Roger Pau Monné wrote:
>> >On Thu, Sep 21, 2017 at 11:01:56PM -0400, Lan Tianyu wrote:
>> >> +sta
On Fri, Oct 20, 2017 at 12:20:06PM +0100, Roger Pau Monné wrote:
>On Thu, Sep 21, 2017 at 11:02:09PM -0400, Lan Tianyu wrote:
>> From: Chao Gao
>>
>> Queued Invalidation Interface is an expanded invalidation interface with
>> extended capabilities. Hardware impleme
On Mon, Oct 23, 2017 at 09:57:16AM +0100, Roger Pau Monné wrote:
>On Mon, Oct 23, 2017 at 03:50:24PM +0800, Chao Gao wrote:
>> On Fri, Oct 20, 2017 at 12:20:06PM +0100, Roger Pau Monné wrote:
>> >On Thu, Sep 21, 2017 at 11:02:09PM -0400, Lan Tianyu wrote:
&g
destroying domain.
Signed-off-by: Lan Tianyu
Signed-off-by: Chao Gao
---
v4:
- introduce REGISTER_VIOMMU() to register viommu types and ops.
- remove unneeded domctl interface to destroy viommu.
---
docs/misc/xen-command-line.markdown | 7 ++
xen/arch/x86/Kconfig| 1 +
xen/arch
From: Lan Tianyu
This patch is to add Xen virtual IOMMU doc to introduce motivation,
framework, vIOMMU hypercall and xl configuration.
Signed-off-by: Lan Tianyu
Signed-off-by: Chao Gao
---
docs/misc/viommu.txt | 120 +++
1 file changed, 120
From: Lan Tianyu
This patch is to add callback for vIOAPIC and vMSI to check whether interrupt
remapping is enabled.
Signed-off-by: Lan Tianyu
Signed-off-by: Chao Gao
---
xen/common/viommu.c | 15 +++
xen/include/xen/viommu.h | 4
2 files changed, 19 insertions
From: Lan Tianyu
This patch is to add get_irq_info callback for platform implementation
to convert irq remapping request to irq info (E,G vector, dest, dest_mode
and so on).
Signed-off-by: Lan Tianyu
Signed-off-by: Chao Gao
---
xen/common/viommu.c | 16
xen/include
Intel VTD).
5) Add save and restore support for vvtd
Chao Gao (23):
vtd: clean-up and preparation for vvtd
x86/hvm: Introduce a emulated VTD for HVM
x86/vvtd: Add MMIO handler for VVTD
x86/vvtd: Set Interrupt Remapping Table Pointer through GCMD
x86/vvtd: Enable Interrupt Remappi
From: Lan Tianyu
This patch is to add irq request callback for platform implementation
to deal with irq remapping request.
Signed-off-by: Lan Tianyu
Signed-off-by: Chao Gao
---
xen/common/viommu.c | 15
xen/include/asm-x86/viommu.h | 54
track
info (e.g. the table's gfn and max supported entries) of interrupt remapping
table.
Signed-off-by: Chao Gao
Signed-off-by: Lan Tianyu
---
v4:
- declare eim_enabled as bool and irt as gfn_t
- rename vvtd_handle_gcmd_sirtp() to write_gcmd_sirtp()
v3:
- ignore unaligned r/w of vt-d har
This patch adds create/destroy function for the emulated VTD
and adapts it to the common VIOMMU abstraction.
As the Makefile is changed here, put all files in alphabetic order
by this chance.
Signed-off-by: Chao Gao
Signed-off-by: Lan Tianyu
---
v4:
- use REGISTER_VIOMMU
- shrink the size of
g
flags to show that some operations are completed. These operations
include enabling interrupt remapping and performing a kind of invalidation
requests. In vvtd, we also try to drain in-flight interrupts by waiting
the inflight_intr is decreased to 0.
Signed-off-by: Chao Gao
Signed-off-by: Lan Tiany
.
Signed-off-by: Chao Gao
Signed-off-by: Lan Tianyu
---
v4:
- Only fix the alignment and defer introducing new definition to when
they are needed
(Suggested-by Roger Pau Monné)
- remove parts of open-coded masks
v3:
- new
---
xen/drivers/passthrough/vtd/iommu.h | 86
ff-by: Chao Gao
Signed-off-by: Lan Tianyu
---
v4:
- map guest's interrupt reampping table to Xen permanently rather than
mapping one specific page on demand.
---
xen/drivers/passthrough/vtd/iommu.h | 3 +-
xen/drivers/passthrough/vtd/vvtd.c | 98 +
2 fil
When IOAPIC RTE is in remapping format, it doesn't contain the vector of
interrupt. For this case, the RTE contains an index of interrupt remapping
table where the vector of interrupt is stored. This patchs gets the vector
through a vIOMMU interface.
Signed-off-by: Chao Gao
Signed-off-by
buffer in system memory. Software submits commands by writing Invalidation
Descriptors to the IQ.
In this patch, a new function viommu_process_iq() is used for emulating how
hardware handles invalidation requests through QI.
Signed-off-by: Chao Gao
Signed-off-by: Lan Tianyu
---
v4
Provide a save-restore pair to save/restore registers and non-register
status.
Signed-off-by: Chao Gao
Signed-off-by: Lan Tianyu
---
v3:
- use one entry to save both vvtd registers and other intermediate
state
---
xen/drivers/passthrough/vtd/vvtd.c | 57
. vioapic_get_vector(). With vIOMMU, the RTE may don't contain vector.
2. perform EOI which is always based on the interrupt vector.
Signed-off-by: Chao Gao
Signed-off-by: Lan Tianyu
---
v3:
- add example cases in which we will use this function.
---
xen/drivers/passthrough/vtd/vvtd.c
A field, viommu_info, is added to struct libxl_domain_build_info. Several
attributes can be specified by guest config file for virtual IOMMU. These
attributes are used for DMAR construction and vIOMMU creation.
Signed-off-by: Chao Gao
Signed-off-by: Lan Tianyu
---
v4:
- Move
process_iq() in vvtd_read() rather than in
vvtd_handle_irq_request() is to avoid ABBA deadlock of d->event_lock and
vvtd->ie_lock.
Signed-off-by: Chao Gao
Signed-off-by: Lan Tianyu
---
v4:
- new
---
xen/arch/x86/hvm/hvm.c | 2 +-
xen/drivers/passth
re used to search the physical irq. It is clear that for
remapping format msi, the interrupt attributs should be decodes from
IRTE.
Signed-off-by: Chao Gao
Signed-off-by: Lan Tianyu
---
xen/arch/x86/hvm/irq.c | 6 ++
xen/arch/x86/hvm/vmsi.c | 33 +---
Add dmar table structure according Chapter 8 "BIOS Considerations" of
VTd spec Rev. 2.4.
VTd
spec:http://www.intel.com/content/dam/www/public/us/en/documents/product-specifications/vt-directed-io-spec.pdf
Signed-off-by: Chao Gao
Signed-off-by: Lan Tianyu
---
tools/libacpi/acpi
Software writes to QIE field of GCMD to enable or disable queued
invalidations. This patch emulates QIE field of GCMD.
Signed-off-by: Chao Gao
Signed-off-by: Lan Tianyu
---
xen/drivers/passthrough/vtd/iommu.h | 3 ++-
xen/drivers/passthrough/vtd/vvtd.c | 18 ++
2 files
A new logic is added to init_acpi_config(). The logic initializes
some fields introduced for DMAR table. For PVH guest, the DMAR table
is built as other tables. But for HVM guest, only the DMAR table is
built in toolstack and pass through it to guest via existing mechanism.
Signed-off-by: Chao
When irq remapping is enabled, IOAPIC Redirection Entry may be in remapping
format. If that, generate an irq_remapping_request and call the common
VIOMMU abstraction's callback to handle this interrupt request. Device
model is responsible for checking the request's validity.
Signed-of
_fault() reports faults to software.
Currently, only Primary Fault Logging is supported and the Number of
Fault-recording Registers is 1.
Signed-off-by: Chao Gao
Signed-off-by: Lan Tianyu
---
v4:
- introduce a lock to protect fault-event related regs
---
xen/drivers/passthrough/vtd/iommu.h |
significantly reduce the efforts to support binding
remappable format msi.
Signed-off-by: Chao Gao
Signed-off-by: Lan Tianyu
---
v4:
- new
---
tools/libxc/include/xenctrl.h | 7 ---
tools/libxc/xc_domain.c | 14 --
xen/arch/x86/hvm/vmsi.c | 12 ++--
xen/drivers
.
Signed-off-by: Chao Gao
Signed-off-by: Lan Tianyu
---
v3:
- new
---
xen/drivers/passthrough/vtd/vvtd.c | 10 ++
1 file changed, 10 insertions(+)
diff --git a/xen/drivers/passthrough/vtd/vvtd.c
b/xen/drivers/passthrough/vtd/vvtd.c
index 9890cc2..d3dec01 100644
--- a/xen/drivers
This patch adds VVTD MMIO handler to deal with MMIO access.
Signed-off-by: Chao Gao
Signed-off-by: Lan Tianyu
---
v4:
- only trap the register emulated in vvtd_in_range().
i.e. replace PAGE_SIZE with the VVTD_MAX_OFFSET
---
xen/drivers/passthrough/vtd/vvtd.c | 55
a pirq.
Signed-off-by: Chao Gao
Signed-off-by: Lan Tianyu
---
v3:
- clarify the interrupt format bit is Intel-specific, then it is
improper to define MSI_ADDR_IF_MASK in a common header.
---
hw/i386/xen/xen-hvm.c | 10 +-
hw/pci/msi.c | 5 +++--
hw/pci/msix.c | 4 +++-
If guest is configured to have a vIOMMU, create it during domain construction.
Signed-off-by: Chao Gao
Signed-off-by: Lan Tianyu
---
v4:
- s/LOGED/LOGD
v3:
- Remove the process of querying capabilities.
---
tools/libxl/libxl_x86.c | 17 +
1 file changed, 17 insertions
exist interface is modified to support msi of both format.
- In patch 3, define MSI_ADDR_IF_MASK inside a function because
it is intel-specific. It is improper to define it in a common header.
Chao Gao (3):
i386/msi: Correct mask of destination ID in MSI address
xen/pt: Pass the whole msi addr/data
Add libxc helpers for XEN_DOMCTL_viommu_op. Now, it has one sub-command
- create(): create a vIOMMU in Xen, given viommu type, register-set
location and capabilities
Signed-off-by: Chao Gao
Signed-off-by: Lan Tianyu
---
v4:
- remove destroy() sub-command
v3:
- Remove API for
table according to the new fields.
The header files in ovmf.c are re-ordered to avoid including in
tools/libacpi/libacpi.h.
Signed-off-by: Chao Gao
Signed-off-by: Lan Tianyu
---
v4:
- initialize variables during declaration if possible
- reorder the sequence of header files to avoid including
filtering.
The main reason why we want this is QEMU doesn't have the knowledge
to decide the interrupt format after we introduce vIOMMU inside Xen.
Passing the whole msi message down and let arch-specific vIOMMU to
decide the interrupt format.
Signed-off-by: Chao Gao
Signed-off-by: Lan Tianyu
-
0x1ef/0x42c
(XEN)[] entry.o#test_all_events+0/0x30
(XEN)
(XEN)
(XEN)
(XEN) Panic on CPU 2:
(XEN) Xen BUG at spinlock.c:47
(XEN)
(XEN)
(XEN) Reboot in five seconds...
Signed-off-by: Chao Gao
---
xen/drivers/passth
On Fri, Jul 21, 2017 at 05:26:47PM +0100, George Dunlap wrote:
>On Fri, Jul 7, 2017 at 7:49 AM, Chao Gao wrote:
>> In order to analyze PI blocking list operation frequence and obtain
>> the list length, add some relevant events to xentrace and some
>> associated co
Hi, everyone.
I have a machine which has two numa nodes. The NODE0 contains
memory range from 0 to 0x18400MB and NODE1 contains memory range
from 0x18400MB to 0x1c400MB. The resource available to dom0 is
restricted through adding "dom0_mem=10G dom0_nodes=0 dom0_max_vcpus=48"
to Xen Command Line. E
On Mon, Jun 19, 2017 at 01:43:25AM -0600, Jan Beulich wrote:
On 19.06.17 at 08:33, wrote:
>> On Fri, Jun 16, 2017 at 09:52:11AM -0600, Jan Beulich wrote:
>> On 16.06.17 at 08:48, wrote:
The problem is a VF of RC integrated PF (e.g. PF's BDF is 00:02.0),
we would wrongly use 00:
I-capable but may have devfn > 7.
This patch directly looks up the 'is_extfn' field of PF's struct pci_dev
to decide whether the PF is a extended function.
Reported-by: Crawford Eric R
Signed-off-by: Chao Gao
---
xen/drivers/passthrough/vtd/dmar.c | 12 +++-
1 file chan
On Fri, Jun 16, 2017 at 08:34:20AM -0600, Jan Beulich wrote:
On 24.05.17 at 08:56, wrote:
>> --- a/xen/arch/x86/hvm/vmx/vmx.c
>> +++ b/xen/arch/x86/hvm/vmx/vmx.c
>> @@ -82,6 +82,7 @@ static int vmx_vmfunc_intercept(struct cpu_user_regs
>> *regs);
>> struct vmx_pi_blocking_vcpu {
>> str
On Thu, Jun 22, 2017 at 03:26:04AM -0600, Jan Beulich wrote:
On 21.06.17 at 12:47, wrote:
>> The problem is a VF of RC integrated PF (e.g. PF's BDF is 00:02.0),
>> we would wrongly use 00:00.0 to search VT-d unit.
>>
>> To search VT-d unit for a VF, the BDF of the PF is used. And If the
>> P
On Fri, Jun 16, 2017 at 09:09:13AM -0600, Jan Beulich wrote:
On 24.05.17 at 08:56, wrote:
>> Currently, a blocked vCPU is put in its pCPU's pi blocking list. If
>> too many vCPUs are blocked on a given pCPU, it will incur that the list
>> grows too long. After a simple analysis, there are 32k
On Fri, Jun 23, 2017 at 01:58:52AM -0600, Jan Beulich wrote:
On 23.06.17 at 06:22, wrote:
>> On Fri, Jun 16, 2017 at 09:09:13AM -0600, Jan Beulich wrote:
>> On 24.05.17 at 08:56, wrote:
+{
+pi_cpu = cpumask_cycle(pi_cpu, &cpu_online_map);
>>>
>>>With this, how could
e original code did; The original code wasn't aware the former.
This patch directly looks up the 'is_extfn' field of PF's struct pci_dev
to decide whether the PF is a extended function.
Reported-by: Crawford, Eric R
Signed-off-by: Chao Gao
---
xen/drivers/passthrough/vtd/d
On Thu, Jun 29, 2017 at 04:42:33PM +0100, Roger Pau Monné wrote:
>On Thu, Jun 29, 2017 at 11:21:53AM +0800, Chao Gao wrote:
>> The problem is for a VF of RC integrated PF (e.g. PF's BDF is 00:02.0),
>> we would wrongly use 00:00.0 to search VT-d unit.
>>
>> From
e original code did; The original code wasn't aware the former.
This patch directly looks up the 'is_extfn' field of PF's struct pci_dev
to decide whether the PF is a extended function.
Reported-by: Crawford, Eric R
Signed-off-by: Chao Gao
---
v3:
- access pf's struct pci
On Fri, Jun 30, 2017 at 05:19:52PM +0800, Tian, Kevin wrote:
>> From: Gao, Chao
>> Sent: Friday, June 30, 2017 9:17 AM
>>
>> The problem is for a VF of RC integrated PF (e.g. PF's BDF is 00:02.0),
>> we would wrongly use 00:00.0 to search VT-d unit.
>>
>> From SRIOV spec REV 1.0 section 3.7.3, it
On Wed, Jul 05, 2017 at 10:46:39AM +0800, Tian, Kevin wrote:
>> From: Gao, Chao
>> Sent: Monday, July 3, 2017 12:37 PM
>>
>> On Fri, Jun 30, 2017 at 05:19:52PM +0800, Tian, Kevin wrote:
>> >> From: Gao, Chao
>> >> Sent: Friday, June 30, 2017 9:17 AM
>> >>
>> >> The problem is for a VF of RC integr
On Wed, Jul 05, 2017 at 01:18:38PM +0800, Tian, Kevin wrote:
>> From: Gao, Chao
>> Sent: Wednesday, July 5, 2017 12:28 PM
>>
>> On Wed, Jul 05, 2017 at 10:46:39AM +0800, Tian, Kevin wrote:
>> >> From: Gao, Chao
>> >> Sent: Monday, July 3, 2017 12:37 PM
>> >>
>> >> On Fri, Jun 30, 2017 at 05:19:52P
On Wed, Jul 05, 2017 at 09:29:29AM +0100, Roger Pau Monné wrote:
>On Wed, Jul 05, 2017 at 02:19:17AM -0600, Jan Beulich wrote:
>> >>> On 05.07.17 at 09:56, wrote:
>> > How about changing the second paragraph to:
>> >
>> > If a PF is an extended function, the BDF of a traditional function
>> > wit
ing 'is_extfn' field of struct pci_dev has been passed down from
Domain0 to indicate whether the function is an extended function, this patch
just looks up the 'is_extfn' field of PF's struct pci_dev and set 'devfn' to 0
when 'is_extfn' is true.
Repo
This patch adds a field, counter, in struct vmx_pi_blocking_vcpu to track
how many entries are on the pi blocking list.
Signed-off-by: Chao Gao
---
v4:
- non-trace part of Patch 1 in v3
---
xen/arch/x86/hvm/vmx/vmx.c | 14 +++---
1 file changed, 11 insertions(+), 3 deletions(-)
diff
20The%20length%20of%20the%20list%20depends;#422567
[3]
https://lists.gt.net/xen/devel/472749?search_string=enable%20vt-d%20pi%20by%20default;#472749
Chao Gao (4):
VT-d PI: track the vcpu number on pi blocking list
x86/vcpu: track hvm vcpu number on the system
VT-d PI: restrict the vcpu number o
This number is used to calculate the average vcpus per pcpu ratio.
Signed-off-by: Chao Gao
---
v4:
- move the place we increase/decrease the hvm vcpu number to
hvm_vcpu_{initialise, destory}
---
xen/arch/x86/hvm/hvm.c| 6 ++
xen/include/asm-x86/hvm/hvm.h | 3 +++
2 files changed
th of the pcpu where the vcpu is running exceeds
the limit which is the average vcpus per pcpu ratio plus a constant.
If no, the vcpu is added to this pcpu's pi blocking list. Otherwise,
another online pcpu is chosen to accept the vcpu.
Signed-off-by: Chao Gao
---
v4:
- use a new lock to avoid
suitable any more. Thus, this patch extends the toplevel_assert_check()
to remove such assumptions for events of type ASYNC_PI_LIST_DEL.
Signed-off-by: Chao Gao
---
v4:
- trace part of Patch 1 in v3
---
tools/xentrace/formats | 3 +
tools/xentrace/xenalyze.c | 154
On Fri, Jul 07, 2017 at 09:37:04AM -0600, Jan Beulich wrote:
On 07.07.17 at 08:49, wrote:
>> --- a/xen/arch/x86/hvm/vmx/vmx.c
>> +++ b/xen/arch/x86/hvm/vmx/vmx.c
>> @@ -190,7 +190,9 @@ static void vmx_vcpu_block(struct vcpu *v)
>> */
>> ASSERT(old_lock == NULL);
>>
>> -per_cp
On Fri, Jul 07, 2017 at 09:41:18AM -0600, Jan Beulich wrote:
>>>> On 07.07.17 at 08:48, wrote:
>> This patch adds a field, counter, in struct vmx_pi_blocking_vcpu to track
>> how many entries are on the pi blocking list.
>>
>> Signed-off-by: Chao Gao
&g
On Fri, Jul 07, 2017 at 09:57:47AM -0600, Jan Beulich wrote:
On 07.07.17 at 08:48, wrote:
>> --- a/xen/arch/x86/hvm/vmx/vmx.c
>> +++ b/xen/arch/x86/hvm/vmx/vmx.c
>> @@ -95,22 +95,91 @@ static DEFINE_PER_CPU(struct vmx_pi_blocking_vcpu,
>> vmx_pi_blocking);
>> uint8_t __read_mostly posted_in
On Mon, Jul 10, 2017 at 03:36:52AM -0600, Jan Beulich wrote:
On 10.07.17 at 03:17, wrote:
>> On Fri, Jul 07, 2017 at 09:57:47AM -0600, Jan Beulich wrote:
>> On 07.07.17 at 08:48, wrote:
+#define remote_pbl_operation_begin(flags) \
+({
On Mon, Sep 11, 2017 at 10:14:15AM +0100, Roger Pau Monné wrote:
>On Mon, Sep 11, 2017 at 09:58:01AM +0800, Chao Gao wrote:
>> Hi, Roger.
>>
>> I meet an error when creating a pvh guest. I am using commit 6e2a4c73564a.
>> From the error log, I found bootlate_pv()-&g
On Wed, Mar 29, 2017 at 03:28:43AM +, Xuquan (Quan Xu) wrote:
>On March 22, 2017 2:14 PM, Chao Gao wrote:
>>On Wed, Mar 22, 2017 at 06:47:33AM -0600, Jan Beulich wrote:
>>>> 3. We read RTE 3 times. 1st happens when we set vIRR. 2nd happens
>>>> when
>>>
Tianyu is on vacation this two weeks, so I will try to address
some comments on this series.
On Tue, Mar 28, 2017 at 05:24:03PM +0100, Wei Liu wrote:
>On Fri, Mar 17, 2017 at 07:27:05PM +0800, Lan Tianyu wrote:
>> From: Chao Gao
>>
>> In previous patch, we introduce a co
On Wed, Mar 29, 2017 at 09:00:58AM +0100, Roger Pau Monné wrote:
>On Tue, Mar 21, 2017 at 01:29:26PM +0800, Lan Tianyu wrote:
>> On 2017年03月21日 10:28, Lan Tianyu wrote:
>> > On 2017年03月20日 22:23, Roger Pau Monné wrote:
>> >> Thanks! So you add all this vIOMMU code, but the maximum number of allowed
From: Feng Wu
Use type-safe structure assignment instead of memcpy()
Use sizeof(*iremap_entry).
Signed-off-by: Feng Wu
Signed-off-by: Chao Gao
Reviewed-by: Konrad Rzeszutek Wilk
Acked-by: Kevin Tian
---
v10:
- Added several lines for patch [1/6] has been reworked.
v9:
- Delete several
remap_entry() to update IRTE to posted format.
Signed-off-by: Feng Wu
Signed-off-by: Chao Gao
---
v11:
- Commit message changes.
- Add code to clear the two new fields introduced in this patch.
v10:
- Newly added.
xen/arch/x86/msi.c | 1 +
xen/drivers/passthrough/io.c
From: Feng Wu
When cpu is offline, we need to move all the vcpus in its blocking
list to another online cpu, this patch handles it.
Signed-off-by: Feng Wu
Signed-off-by: Chao Gao
Reviewed-by: Jan Beulich
Acked-by: Kevin Tian
---
xen/arch/x86/hvm/vmx/vmcs.c | 1 +
xen/arch/x86/hvm
migration so pirq directly injected to guest needn't be migrated.
This patch adds a indicator, @posted, to show whether the pt irq is delivered
through VT-d PI.
Signed-off-by: Chao Gao
---
v11:
- rename the indicator to 'posted'
- move setting 'posted' field to event lo
when guest wrongly configurate destination of pirq or makes it have
multi-destination vcpus.
Signed-off-by: Chao Gao
Reviewed-by: Jan Beulich
Reviewed-by: Kevin Tian
---
v11:
- move the code (one line) that allow the parameter 'vcpu' of pi_update_irte()
can be NULL to Patch [2/6].
v10
irte. Both variants will update IRTE if possible. If the caller requests a
atomic update but we can't meet it, we raise a bug.
Signed-off-by: Chao Gao
---
v11:
- Add two variant function to update IRTE. Call the non-atomic one for init
and clear operations. Call the atomic one for other
101 - 200 of 294 matches
Mail list logo
