On 09/04/2015 10:18, Chao Peng wrote:
> For each socket, a COS to CBM mapping structure is maintained for each
> COS. The mapping is indexed by COS and the value is the corresponding
> CBM. Different VMs may use the same CBM, a reference count is used to
> indicate if the CBM is available.
>
> Sign
On 09/04/2015 10:18, Chao Peng wrote:
> Detect Intel Cache Allocation Technology(CAT) feature and store the
> cpuid information for later use. Currently only L3 cache allocation is
> supported. The L3 CAT features may vary among sockets so per-socket
> feature information is stored. The initializat
On 09/04/2015 10:18, Chao Peng wrote:
> Switching RMID from previous vcpu to next vcpu only needs to write
> MSR_IA32_PSR_ASSOC once. Write it with the value of next vcpu is enough,
> no need to write '0' first. Idle domain has RMID set to 0 and because MSR
> is already updated lazily, so just swit
On 09/04/2015 21:38, Boris Ostrovsky wrote:
> In case dom0_max_vcpus is incorrectly specified on boot line make sure
> we will still boot.
>
> Signed-off-by: Boris Ostrovsky
Good catch - lets not do that.
Reviewed-by: Andrew Cooper
> ---
> xen/arch/x86/domain_build.c | 2 ++
> 1 file changed,
In case dom0_max_vcpus is incorrectly specified on boot line make sure
we will still boot.
Signed-off-by: Boris Ostrovsky
---
xen/arch/x86/domain_build.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/xen/arch/x86/domain_build.c b/xen/arch/x86/domain_build.c
index e5c845c..378e650 100644
On 09/04/2015 10:18, Chao Peng wrote:
> Change type of opt_psr from bool to int so more psr features can fit.
>
> Introduce a new routine to parse bool parameter so that both cmt and
> future psr features like cat can use it.
>
> Signed-off-by: Chao Peng
Reviewed-by: Andrew Cooper
_
On 04/09/2015 02:23 PM, Peter Zijlstra wrote:
On Thu, Apr 09, 2015 at 08:13:27PM +0200, Peter Zijlstra wrote:
On Mon, Apr 06, 2015 at 10:55:44PM -0400, Waiman Long wrote:
+#define PV_HB_PER_LINE (SMP_CACHE_BYTES / sizeof(struct pv_hash_bucket))
+static struct qspinlock **pv_hash(struct qspinloc
On Thu, Apr 09, 2015 at 09:57:21PM +0200, Peter Zijlstra wrote:
> On Mon, Apr 06, 2015 at 10:55:48PM -0400, Waiman Long wrote:
>
> > @@ -219,24 +236,30 @@ static void pv_wait_node(struct mcs_spinlock *node)
> > }
> >
> > /*
> > + * Called after setting next->locked = 1 & lock acquired.
> > + *
In XenServer, we have a single server which occasionally finds interrupts
unexpectedtly in the wrong state (I suspect the SMM handler). However,
asserts from ASSERT_INTERRUPTS_{EN,DIS}ABLED are quite opaque, being an
unqualified fatal #UD.
Andrew Cooper (3):
x86/bug: Infrastructure to create BU
---
xen/arch/x86/traps.c| 66 +++
xen/arch/x86/x86_64/entry.S | 26 +
2 files changed, 92 insertions(+)
diff --git a/xen/arch/x86/traps.c b/xen/arch/x86/traps.c
index 22cdfc4..c562842 100644
--- a/xen/arch/x86/traps.c
+++ b/xen/a
Signed-off-by: Andrew Cooper
CC: Keir Fraser
CC: Jan Beulich
---
xen/include/asm-x86/asm_defns.h | 25 -
1 file changed, 16 insertions(+), 9 deletions(-)
diff --git a/xen/include/asm-x86/asm_defns.h b/xen/include/asm-x86/asm_defns.h
index 1674c7c..e8a678e 100644
--- a
Signed-off-by: Andrew Cooper
CC: Keir Fraser
CC: Jan Beulich
---
xen/include/asm-x86/bug.h | 48 -
1 file changed, 43 insertions(+), 5 deletions(-)
diff --git a/xen/include/asm-x86/bug.h b/xen/include/asm-x86/bug.h
index cd862e3..365c6b8 100644
---
On Mon, Apr 06, 2015 at 10:55:48PM -0400, Waiman Long wrote:
> @@ -219,24 +236,30 @@ static void pv_wait_node(struct mcs_spinlock *node)
> }
>
> /*
> + * Called after setting next->locked = 1 & lock acquired.
> + * Check if the the CPU has been halted. If so, set the _Q_SLOW_VAL flag
> + * and
On Thu, Apr 9, 2015 at 8:29 PM, George Dunlap
wrote:
> Add core functionality and an executable to run it
>
> Signed-off-by: George Dunlap
> ---
> CC: Stefano Stabellini
Hmm, not sure where the 0/8 for this patch went... anyway, you can
also find this series here:
git://xenbits.xenproject.org/
First, create a new global variable, PKGTYPE. At the moment support "deb" and
"rpm".
Define _check-package-$PKGTYPE which returns true if the package is
installed, false otherwise, and _install-package-$PKGTYPE which will
install a list of packages.
Define check-package(), which will take a lis
Add core functionality and an executable to run it
Signed-off-by: George Dunlap
---
CC: Stefano Stabellini
---
lib/core.sh | 147
raise | 16 +++
2 files changed, 163 insertions(+)
create mode 100644 lib/core.sh
create mo
Make as few changes as possible to begin with, just to separate code
motion from changes.
For now, remove raise.sh and unraise.sh from package creation, until
we can figure out what to do instead.
Signed-off-by: George Dunlap
---
CC: Stefano Stabellini
---
raise.sh => lib/build.sh | 62 +++
Update to the core.sh calling convention, which will parse var=value
arguments.
Also tweak some Fedora dependencies and refactor the build dependency
checking / installation code.
CC: Stefano Stabellini
___
Xen-devel mailing list
Xen-devel@lists.xen.
These should be called from other places that have already done the
appropriate "source"
Signed-off-by: George Dunlap
---
CC: Stefano Stabellini
---
components/grub| 4
components/libvirt | 4
components/xen | 4
3 files changed, 12 deletions(-)
diff --git a/components/g
The xen build requires eiter wget or ftp. Use wget.
Signed-off-by: George Dunlap
---
CC: Stefano Stabellini
---
components/xen | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/components/xen b/components/xen
index e8cbb2b..7a9f22d 100644
--- a/components/xen
+++ b/compon
acpica-tools is the new package that replaces iasl.
Signed-off-by: George Dunlap
---
CC: Stefano Stabellini
---
components/xen | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/components/xen b/components/xen
index 70b72b0..e8cbb2b 100644
--- a/components/xen
+++ b/components/
"script" implies something which is designed to be run standalone.
"lib" implies that this is going to be sourced from another bash script.
Also change "git-checkout" to be a function rather than a script
Signed-off-by: George Dunlap
---
CC: Stefano Stabellini
---
components/grub
Move the default config to "defconfig" and add config to .gitignore.
If config doesn't exist, then copy defconfig to it in common_init.
Signed-off-by: George Dunlap
---
CC: Stefano Stabellini
---
.gitignore | 1 +
config => defconfig | 0
raise | 13 -
3 fil
Wei Liu writes ("[PATCH v4 2/3] libxl: use new QEMU xenstore protocol"):
> Originally both QEMU traditional and QEMU upstream used hardcoded
> /local/domain/0 paths. This patch changes the protocol to use
> /local/domain/$dm_domid path.
Thanks. I have committed 1/2 and 2/2 (the latter with the ap
For QEMU traditional running in Dom0, there is no functional change
because it will still write to the same /local/domain/0 path.
For QEMU traditional stubdom, this is an incompatible startup
protocol change. There is a corresponding libxl changeset "libxl:
use new QEMU xenstore protocol", which
Originally both QEMU traditional and QEMU upstream used hardcoded
/local/domain/0 paths. This patch changes the protocol to use
/local/domain/$dm_domid path.
For QEMU traditional and upstream without stubdom, $dm_domid is 0 so
the path is in fact still /local/domain/0.
For QEMU traditional stubdo
This patch sereis contains 1 patch to QEMU traditional tree and 2 patches to
Xen tree.
Note that QEMU trad patch and libxl patch need to be committed at the same
time. And please fold in the change to update Config.mk to libxl patch to
avoid regression.
The third patch has not been acked, but th
This reverts commit dd748d128d86996592afafea02e578cc7d4e6d42.
We don't need this workaround anymore since we have fixed the toolstack
interlock problem that affects stubdom.
Signed-off-by: Wei Liu
Cc: Paul Durrant
Cc: Jan Beulich
Acked-by: Ian Campbell
---
xen/arch/x86/hvm/hvm.c |
On 31 March 2015 at 17:31, Mark Chambers wrote:
>
> On 31 March 2015 at 11:56, Mark Chambers wrote:
>>
>>
>> It's nested under Hyper-V in the same manner as the problematic install.
>> I was deliberately trying to replicate the issue, but the problem doesn't
>> manifest.
>>
>> Mark
>>
>>>
>>
> H
On Thu, 2015-04-09 at 18:25 +0100, Ian Jackson wrote:
> root@bedbug:~# ethtool -S eth0 | grep -v ': 0$'
> NIC statistics:
> rx_octets: 8196868
> rx_ucast_packets: 633
> rx_mcast_packets: 1
> rx_bcast_packets: 123789
> tx_octets: 42854
> tx_ucast_packets: 9
> tx_m
On Thu, 2015-04-09 at 18:25 +0100, Ian Jackson wrote:
> root@bedbug:~# ethtool -S eth0 | grep -v ': 0$'
> NIC statistics:
> rx_octets: 8196868
> rx_ucast_packets: 633
> rx_mcast_packets: 1
> rx_bcast_packets: 123789
> tx_octets: 42854
> tx_ucast_packets: 9
> tx_mc
On Thu, Apr 09, 2015 at 08:13:27PM +0200, Peter Zijlstra wrote:
> On Mon, Apr 06, 2015 at 10:55:44PM -0400, Waiman Long wrote:
> > +#define PV_HB_PER_LINE (SMP_CACHE_BYTES / sizeof(struct
> > pv_hash_bucket))
> > +static struct qspinlock **pv_hash(struct qspinlock *lock, struct pv_node
> > *
On Mon, Apr 06, 2015 at 10:55:44PM -0400, Waiman Long wrote:
> +++ b/kernel/locking/qspinlock_paravirt.h
> @@ -0,0 +1,321 @@
> +#ifndef _GEN_PV_LOCK_SLOWPATH
> +#error "do not include this file"
> +#endif
> +
> +/*
> + * Implement paravirt qspinlocks; the general idea is to halt the vcpus
> instea
Wei Liu writes ("Re: [PATCH] qemu-trad: xenstore: use relative path for
device-model node"):
> On Thu, Apr 09, 2015 at 06:46:31PM +0100, Ian Jackson wrote:
> > Right. So that means that this patch needs to go in at the same time
> > as the corresponding libxl change.
>
> I don't follow "go in at
On Thu, Apr 09, 2015 at 06:46:31PM +0100, Ian Jackson wrote:
> Wei Liu writes ("Re: [PATCH] qemu-trad: xenstore: use relative path for
> device-model node"):
> > On Thu, Apr 09, 2015 at 05:47:08PM +0100, Ian Jackson wrote:
> > >
> > > I think you mean:
> > > ...
> >
> > So far so good.
> >
> >
Wei Liu writes ("Re: [PATCH] qemu-trad: xenstore: use relative path for
device-model node"):
> On Thu, Apr 09, 2015 at 05:47:08PM +0100, Ian Jackson wrote:
> >
> > I think you mean:
> > ...
>
> So far so good.
>
> > QEMU traditional stubdom was broken by and is still broken in
> > -uns
At 16:34 +0100 on 09 Apr (1428597298), Andrew Cooper wrote:
> On 09/04/15 16:15, Tim Deegan wrote:
> > At 18:26 +0100 on 07 Apr (1428431180), Andrew Cooper wrote:
> >> --- a/xen/arch/x86/boot/head.S
> >> +++ b/xen/arch/x86/boot/head.S
> >> @@ -127,7 +127,8 @@ __start:
> >> mov $sym_phy
Ian Jackson writes ("Re: tg3 NIC driver bug in 3.14.x under Xen [and 3 more
messages]"):
> Prashant Sreedharan writes ("Re: tg3 NIC driver bug in 3.14.x under Xen [and
> 3 more messages]"):
> > yes please do
>
> I will do so.
I did this test:
- Linux 3.14.21
- baremetal
- `iommu=soft swiotlb
Julien Grall writes ("Re: [Xen-devel] [PATCH v5 p2 16/19] tools/libxl: arm: Use
an higher value for the GIC phandle"):
> On 09/04/15 17:52, Ian Jackson wrote:
> > What would happen if our assumption about the DT compiler were
> > violated ?
>
> The phandle would be present in 2 different nodes of
On Thu, Apr 09, 2015 at 05:47:08PM +0100, Ian Jackson wrote:
> Wei Liu writes ("[PATCH] qemu-trad: xenstore: use relative path for
> device-model node"):
> > For QEMU traditional stubdom, this is incompatible startup protocol
> > change. This change needs to work with corresponding libxl changese
On Thu, 2015-04-09 at 17:36 +0100, Stefano Stabellini wrote:
> A very big difference:
>
> echo 262144 > /proc/sys/net/ipv4/tcp_limit_output_bytes
> brings us much closer to the original performance, the slowdown is just
> 8%
Cool.
>
> echo 1048576 > /proc/sys/net/ipv4/tcp_limit_output_bytes
>
Prashant Sreedharan writes ("Re: tg3 NIC driver bug in 3.14.x under Xen [and 3
more messages]"):
> On Thu, 2015-04-09 at 12:11 +0100, Ian Jackson wrote:
> > No. I can try to repro the problem without the bridge, if it would
> > help.
> yes please do
I will do so.
FYI, when I came back to this t
On 09/04/15 18:00, Julien Grall wrote:
> On 09/04/15 17:52, Ian Jackson wrote:
>> Julien Grall writes ("Re: [Xen-devel] [PATCH v5 p2 16/19] tools/libxl: arm:
>> Use an higher value for the GIC phandle"):
>>> On 09/04/15 17:17, Ian Jackson wrote:
>>> I only expect people using the partial device tr
On 09/04/15 17:52, Ian Jackson wrote:
> Julien Grall writes ("Re: [Xen-devel] [PATCH v5 p2 16/19] tools/libxl: arm:
> Use an higher value for the GIC phandle"):
>> On 09/04/15 17:17, Ian Jackson wrote:
>> I only expect people using the partial device tree in very specific use
>> case. Generic use
Julien Grall writes ("Re: [Xen-devel] [PATCH v5 p2 16/19] tools/libxl: arm: Use
an higher value for the GIC phandle"):
> On 09/04/15 17:17, Ian Jackson wrote:
> > I have to say I have no idea what a phandle is...
>
> A phandle is a way to reference another node in the device tree.
> Any node that
Euan Harris writes ("Re: [RFC PATCH v2 00/29] libxl: Cancelling asynchronous
operations"):
> Yes, that would work, but an open loop approach like that can lead to
> frustratingly unreliable tests. I think it would be best to make
> the test aware of the state of the helper - or even in control o
Wei Liu writes ("[PATCH] qemu-trad: xenstore: use relative path for
device-model node"):
> For QEMU traditional stubdom, this is incompatible startup protocol
> change. This change needs to work with corresponding libxl changeset.
> QEMU traditional is shipped with Xen so we are allowed to do suc
On Tue, Apr 07, 2015 at 06:19:52PM +0100, Ian Jackson wrote:
> On the contrary, I think many long-running operations, such as suspend
> and migrations, involve multiple iterations of the libxl event loop.
> Actual suspend/migrate is done in a helper process; the main process
> is responsible for pr
On Thu, 9 Apr 2015, Eric Dumazet wrote:
> On Thu, 2015-04-09 at 16:46 +0100, Stefano Stabellini wrote:
> > Hi all,
> >
> > I found a performance regression when running netperf -t TCP_MAERTS from
> > an external host to a Xen VM on ARM64: v3.19 and v4.0-rc4 running in the
> > virtual machine are 3
On Thu, 2015-04-09 at 12:11 +0100, Ian Jackson wrote:
> Prashant Sreedharan writes ("Re: tg3 NIC driver bug in 3.14.x under Xen [and
> 3 more messages]"):
> > On Wed, 2015-04-08 at 14:59 +0100, Ian Jackson wrote:
> > > Ian Jackson writes ("Re: tg3 NIC driver bug in 3.14.x under Xen"):
> > > > The
Hi Ian,
On 09/04/15 17:17, Ian Jackson wrote:
> Julien Grall writes ("[PATCH v5 p2 16/19] tools/libxl: arm: Use an higher
> value for the GIC phandle"):
>> The partial device tree may contains phandle. The Device Tree Compiler
>> tends to allocate the phandle from 1.
>
> I have to say I have no
Julien Grall writes ("[PATCH v5 p2 17/19] libxl: Add support for Device Tree
passthrough"):
> From: Julien Grall
>
> On ARM, every non-PCI device are described in the device tree. Each of
> them can be found via a path.
>
> This patch introduces a very basic support, only the IOMMU will be set
Julien Grall writes ("[PATCH v5 p2 15/19] tools/(lib)xl: Add partial device
tree support for ARM"):
> Let the user to pass additional nodes to the guest device tree. For
> this purpose, everything in the node /passthrough from the partial
> device tree will be copied into the guest device tree.
T
Julien Grall writes ("[PATCH v5 p2 16/19] tools/libxl: arm: Use an higher value
for the GIC phandle"):
> The partial device tree may contains phandle. The Device Tree Compiler
> tends to allocate the phandle from 1.
I have to say I have no idea what a phandle is...
> Reserve the ID 65000 for the
On Thu, 2015-04-09 at 16:46 +0100, Stefano Stabellini wrote:
> Hi all,
>
> I found a performance regression when running netperf -t TCP_MAERTS from
> an external host to a Xen VM on ARM64: v3.19 and v4.0-rc4 running in the
> virtual machine are 30% slower than v3.18.
>
> Through bisection I found
Changes in v20:
* Made xen_pmu_data's domain_id a domid_t type, resulting in
layout/padding changes (patches 2 and 10)
* Added compat checks for xen_pmu_data (patch 2)
* Clarified how vpmu_count is used in a comment (patch 6)
* Moved page management from under vpmu_lock (patch 6)
* Other (mostly
The two routines share most of their logic.
Signed-off-by: Boris Ostrovsky
Reviewed-by: Dietmar Hahn
---
Changes in v20:
* Initialize curr in vpmu_do_msr() at definition time and add 'likely' qualifier
to first 'if' statement.
xen/arch/x86/hvm/vpmu.c| 75 -
Since PMU is now not HVM specific we can move VPMU-related files up from
arch/x86/hvm/ directory.
Specifically:
arch/x86/hvm/vpmu.c -> arch/x86/cpu/vpmu.c
arch/x86/hvm/svm/vpmu.c -> arch/x86/cpu/vpmu_amd.c
arch/x86/hvm/vmx/vpmu_core2.c -> arch/x86/cpu/vpmu_intel.c
include/asm-x86/h
Export Xen's symbols as {} triplet via new XENPF_get_symbol
hypercall
Signed-off-by: Boris Ostrovsky
Acked-by: Daniel De Graaf
Reviewed-by: Konrad Rzeszutek Wilk
Reviewed-by: Dietmar Hahn
Tested-by: Dietmar Hahn
---
xen/arch/x86/platform_hypercall.c | 28 +++
xen/common/sym
Save VPMU state during context switch for both HVM and PV(H) guests.
A subsequent patch ("x86/VPMU: NMI-based VPMU support") will make it possible
for vpmu_switch_to() to call vmx_vmcs_try_enter()->vcpu_pause() which needs
is_running to be correctly set/cleared. To prepare for that, call
context_
Add support for privileged PMU mode (XENPMU_MODE_ALL) which allows privileged
domain (dom0) profile both itself (and the hypervisor) and the guests. While
this mode is on profiling in guests is disabled.
Signed-off-by: Boris Ostrovsky
Acked-by: Jan Beulich
---
xen/arch/x86/hvm/vpmu.c | 40
Move some VPMU initilization operations into __initcalls to avoid performing
same tests and calculations for each vcpu.
Signed-off-by: Boris Ostrovsky
Acked-by: Jan Beulich
---
xen/arch/x86/hvm/svm/vpmu.c | 106 --
xen/arch/x86/hvm/vmx/vpmu_core2.c | 151 ++
Add runtime interface for setting PMU mode and flags. Three main modes are
provided:
* XENPMU_MODE_OFF: PMU is not virtualized
* XENPMU_MODE_SELF: Guests can access PMU MSRs and receive PMU interrupts.
* XENPMU_MODE_HV: Same as XENPMU_MODE_SELF for non-proviledged guests, dom0
can profile itself
Add pmu.h header files, move various macros and structures that will be
shared between hypervisor and PV guests to it.
Move MSR banks out of architectural PMU structures to allow for larger sizes
in the future. The banks are allocated immediately after the context and
PMU structures store offsets
vpmu structure will be used for both HVM and PV guests. Move it from
hvm_vcpu to arch_vcpu.
Signed-off-by: Boris Ostrovsky
Acked-by: Jan Beulich
Reviewed-by: Kevin Tian
Reviewed-by: Dietmar Hahn
Tested-by: Dietmar Hahn
---
xen/include/asm-x86/domain.h | 2 ++
xen/include/asm-x86/hvm/vcpu.h
Add support for handling PMU interrupts for PV(H) guests.
VPMU for the interrupted VCPU is unloaded until the guest issues XENPMU_flush
hypercall. This allows the guest to access PMU MSR values that are stored in
VPMU context which is shared between hypervisor and domain, thus avoiding
traps to hy
With this patch return value of 1 of vpmu_do_msr() will now indicate whether an
error was encountered during MSR processing (instead of stating that the access
was to a VPMU register).
As part of this patch we also check for validity of certain MSR accesses right
when we determine which register i
Code for initializing/tearing down PMU for PV guests
Signed-off-by: Boris Ostrovsky
Acked-by: Kevin Tian
Acked-by: Daniel De Graaf
---
Changes in v20:
* Moved page freeing/unmapping from under vpmu_lock in
pvpmu_init()/pvpmu_finish():
* Using is_hardware_domain() instead of open-coding
* Added
Intercept accesses to PMU MSRs and process them in VPMU module. If vpmu ops
for VCPU are not initialized (which is the case, for example, for PV guests that
are not "VPMU-enlightened") access to MSRs will return failure.
Dump VPMU state for all domains (HVM and PV) when requested.
Signed-off-by:
Hi all,
I found a performance regression when running netperf -t TCP_MAERTS from
an external host to a Xen VM on ARM64: v3.19 and v4.0-rc4 running in the
virtual machine are 30% slower than v3.18.
Through bisection I found that the perf regression is caused by the
prensence of the following commi
2015-04-07 9:10 GMT-04:00 Dario Faggioli :
> On Tue, 2015-04-07 at 11:27 +0100, Andrew Cooper wrote:
>> On 04/04/2015 03:14, Dario Faggioli wrote:
>>
>> > I'm putting here in the cover letter a markdown document I wrote to better
>> > describe my findings and ideas (sorry if it's a bit long! :-D).
Hi Dario,
2015-04-03 22:14 GMT-04:00 Dario Faggioli :
> Hi Everyone,
>
> This RFC series is the outcome of an investigation I've been doing about
> whether we can take better advantage of features like Intel CMT (and of PSR
> features in general). By "take better advantage of" them I mean, for exa
On 09/04/15 16:15, Tim Deegan wrote:
> At 18:26 +0100 on 07 Apr (1428431180), Andrew Cooper wrote:
>> --- a/xen/arch/x86/boot/head.S
>> +++ b/xen/arch/x86/boot/head.S
>> @@ -127,7 +127,8 @@ __start:
>> mov $sym_phys(__bss_end),%ecx
>> sub %edi,%ecx
>> xor %eax
Update FreeBSD guests in OSSTest to FreeBSD 10.1. The following images
should be placed in the osstest images folder:
ftp://ftp.freebsd.org/pub/FreeBSD/releases/VM-IMAGES/10.1-RELEASE/amd64/Latest/FreeBSD-10.1-RELEASE-amd64.raw.xz
ftp://ftp.freebsd.org/pub/FreeBSD/releases/VM-IMAGES/10.1-RELEASE/i
From: Julien Grall
The partial device tree may contains phandle. The Device Tree Compiler
tends to allocate the phandle from 1.
Reserve the ID 65000 for the GIC phandle. I think we can safely assume
that the partial device tree will never contain a such ID.
Signed-off-by: Julien Grall
Acked-by
From: Julien Grall
ARM and x86 use a different hypercall to map an IRQ to a domain.
The hypercall to give IRQ permission to the domain has also been moved
to be an x86 specific function as ARM guest won't be able to manage the IRQ.
We may want to support it later.
Signed-off-by: Julien Grall
C
From: Julien Grall
Let the user to pass additional nodes to the guest device tree. For
this purpose, everything in the node /passthrough from the partial
device tree will be copied into the guest device tree.
The node /aliases will be also copied to allow the user to define
aliases which can be
From: Julien Grall
The functions fdt_{fisrt,next}_subnode may not be available because:
* It has been introduced in 2013 => Doesn't work on Wheezy
* The prototype exists but the functions are not exposed. Don't ask
why...
The later has been fixed recently in the dtc repo [1]
When th
From: Julien Grall
Signed-off-by: Julien Grall
Acked-by: Stefano Stabellini
Acked-by: Ian Campbell
---
Changes in v4:
- Add Ian's ack
Changes in v3:
- Add Stefano's ack
Changes in v2:
- Don't move the call in common code.
---
xen/arch/arm/domctl.c | 11 +
From: Julien Grall
Note that the example is done on Midway whose SMMU driver is not
supported on Xen upstream.
Currently, I don't have other platform where I can test Device Tree
passthrough.
Signed-off-by: Julien Grall
---
Changes in v5:
- Drop references to "non-PCI" in favor of
From: Julien Grall
The option "dtdev" will be used to passthrough a device described
in the device tree to a guest.
Signed-off-by: Julien Grall
Acked-by: Ian Campbell
Cc: Ian Jackson
Cc: Wei Liu
---
Changes in v5:
- Drop "non-PCI" in the commit message
- Add Ian's ack
From: Julien Grall
This is a follow-up of commit 525ee49 "xsm: add device tree labeling
support" which add support for device tree labelling in flask.
Those helpers will be use latter when non-pci passthrough (i.e device
tree) will be added.
Signed-off-by: Julien Grall
Acked-by: Daniel De Graa
From: Julien Grall
On ARM, every non-PCI device are described in the device tree. Each of
them can be found via a path.
This patch introduces a very basic support, only the IOMMU will be set
up correctly. The user will have to:
- Describe the device in the partial device tree
- Map manua
From: Julien Grall
A device node is described by a path. It will be used to retrieved the
node in the device tree and assign the related device to the domain.
Only non-PCI protected by an IOMMU can be assigned to a guest.
Also document the behavior of XEN_DOMCTL_deassign_device in the public
he
At 16:05 +0100 on 09 Apr (1428595536), Andrew Cooper wrote:
> On 09/04/15 16:00, Tim Deegan wrote:
> > At 18:26 +0100 on 07 Apr (1428431176), Andrew Cooper wrote:
> >> --- a/xen/include/asm-x86/numa.h
> >> +++ b/xen/include/asm-x86/numa.h
> >> @@ -9,7 +9,7 @@
> >>
> >> extern int srat_rev;
> >>
Roger Pau Monne writes ("[PATCH v2 1/2] osstest: update FreeBSD guests to
10.1"):
> Update FreeBSD guests in OSSTest to FreeBSD 10.1. The following images
> should be placed in the osstest images folder:
Thanks for the quick turnaround. I have pushed this and 2/2 to
osstest pretest (with my acks
At 18:26 +0100 on 07 Apr (1428431180), Andrew Cooper wrote:
> --- a/xen/arch/x86/boot/head.S
> +++ b/xen/arch/x86/boot/head.S
> @@ -127,7 +127,8 @@ __start:
> mov $sym_phys(__bss_end),%ecx
> sub %edi,%ecx
> xor %eax,%eax
> -rep stosb
> +shr
Remove some unused variables from ts-freebsd-install script. Also make the
third parameter of target_put_guest_image optional and fix both callers of
this function.
Signed-off-by: Roger Pau Monné
Cc: Ian Jackson
---
Osstest/TestSupport.pm | 4 ++--
ts-freebsd-install | 21 ++---
From: Julien Grall
Xen has to release IRQ routed to a domain in order to reuse later.
Currently only SPIs can be routed to the guest so we only need to
browse SPIs for a specific domain.
Furthermore, a guest can crash and leave the IRQ in an incorrect state
(i.e has not been EOIed). Xen will hav
The first patch in this series updates FreeBSD guests in OSSTest to use raw
images instead of qcow2 (which are no longer provided by upstream). The
second patch is a cleanup for ts-freebsd-install which should not change
functionality.
___
Xen-devel m
From: Julien Grall
Introduce spi_to_pending in order retrieve the irq_pending structure for
a specific SPI.
It's not possible to re-use irq_to_pending because it's required a VCPU
and some call of the new function may during domain destruction after
the VCPUs are freed.
Signed-off-by: Julien Gr
From: Julien Grall
On x86, an IRQ is assigned in 2 steps to an HVM guest:
- The toolstack is calling PHYSDEVOP_map_pirq in order to create a
guest PIRQ (IRQ bound to an event channel)
- The emulator (QEMU) is calling DOMCTL_bind_pt_irq in order to
bind the IRQ
On ARM, there is no
From: Julien Grall
Each domain may have a different number of IRQs depending on the devices
assigned to it.
Rather re-using the number of IRQs used by the hardwared GIC, let the
toolstack specify the number of SPIs when the domain is created. This
will avoid to waste memory.
To calculate the nu
From: Julien Grall
This new function will correctly initialize the IOMMU page table for the
current domain.
Also use it in iommu_assign_dt_device even though the current IOMMU
implementation on ARM shares P2M with the processor.
Signed-off-by: Julien Grall
Acked-by: Jan Beulich
Acked-by: Ian
From: Julien Grall
Flask code already provides a helper to copy a string from guest. In a later
patch, the new DT hypercalls will need a similar function.
To avoid code duplication, copy the flask helper (flask_copying_string) to
common code:
- Rename into safe_copy_string_from_guest
- A
Hello all,
This is the fifth version of this patch series to add support for platform
device passthrough on ARM.
In order to passthrough a non-PCI device, the user will have to:
- Map manually MMIO/IRQ
- Describe the device in the newly partial device tree support
- Specify the list o
From: Julien Grall
The toolstack may not have deassigned every device used by a guest.
Therefore we have to go through the device list and remove them before
asking the IOMMU drivers to release memory for this domain.
This can be done by moving the call to the release function when we
relinquish
From: Julien Grall
The maximum size of the copied string has been chosen based on the value
use by XSM in similar case.
Furthermore, Linux seems to allow path up to 4096 characters. Though
this could vary from one OS to another.
Signed-off-by: Julien Grall
---
Changes in v4:
- Dro
From: Julien Grall
Currently, when the device is deassigned from a domain, we directly reassign
to DOM0.
As the device may not have been correctly reset, this may lead to corruption or
expose some part of DOM0 memory. Also, we may have no way to reset some
platform devices.
If Xen reassigns the
On 09/04/15 16:02, Tim Deegan wrote:
> At 18:26 +0100 on 07 Apr (1428431177), Andrew Cooper wrote:
>> Hoist MEMF_node(cpu_to_node(cpu)) to the start of the function, and avoid
>> passing (potentially bogus) memflags if node information is not available.
>>
>> Signed-off-by: Andrew Cooper
> As it h
1 - 100 of 180 matches
Mail list logo
