On Wed, Jun 03, Ian Campbell wrote:
> On Sat, 2015-05-23 at 08:24 +, Olaf Hering wrote:
> > Having xenalyze in the source tree makes it much easier to keep private
> > debug code in hypervisor and xenalyze in sync. It helped alot while
> > debugging the root cause for commit
> > 607e8494c4239
Hi,
On Wed, Jun 3, 2015 at 5:20 PM, Stefano Stabellini <
stefano.stabell...@eu.citrix.com> wrote:
> On Wed, 3 Jun 2015, Sanjeev Pandita wrote:
> > Hi Stefano,
> >
> > On Wed, Jun 3, 2015 at 3:45 PM, Stefano Stabellini <
> stefano.stabell...@eu.citrix.com> wrote:
> > On Wed, 3 Jun 2015, Sanj
flight 58229 rumpuserxen real [real]
http://logs.test-lab.xenproject.org/osstest/logs/58229/
Regressions :-(
Tests which did not succeed and are blocking,
including tests which could not be run:
build-amd64-rumpuserxen 5 rumpuserxen-build fail REGR. vs. 33866
build-i386-rumpuserxe
Em Mon, 08 Jun 2015 17:20:20 -0700
"Luis R. Rodriguez" escreveu:
> From: "Luis R. Rodriguez"
>
> We are burrying direct access to MTRR code support on
> x86 in order to take advantage of PAT. In the future we
> also want to make the default behaviour of ioremap_nocache()
> to use strong UC, use
On 2015/6/7 19:27, Wei Liu wrote:
On Wed, Jun 03, 2015 at 10:58:31AM +0800, Chen, Tiejun wrote:
On 2015/6/3 0:36, Wei Liu wrote:
On Fri, May 22, 2015 at 05:35:08PM +0800, Tiejun Chen wrote:
This patch passes rdm reservation policy to xc_assign_device() so the policy
is checked when assigning d
> -Original Message-
> From: Ian Campbell [mailto:ian.campb...@citrix.com]
> Sent: Monday, June 08, 2015 6:31 PM
> To: Pang, LongtaoX
> Cc: xen-devel@lists.xen.org; ian.jack...@eu.citrix.com; wei.l...@citrix.com;
> Hu,
> Robert
> Subject: Re: [OSSTEST Nested PATCH v11 4/7] Changes on tes
flight 58178 linux-3.4 real [real]
http://logs.test-lab.xenproject.org/osstest/logs/58178/
Regressions :-(
Tests which did not succeed and are blocking,
including tests which could not be run:
test-amd64-amd64-xl 9 debian-install fail REGR. vs. 52209-bisect
test-amd64-amd64-pair
On 06/08/2015 09:04 PM, Andrew Cooper wrote:
On 08/06/15 04:45, Yang Hongyang wrote:
call callbacks->get_dirty_pfn() after suspend primary vm to
get dirty pages on secondary vm, and send pages both dirty on
primary/secondary to secondary.
Signed-off-by: Yang Hongyang
Signed-off-by: Wen Congy
On 06/08/2015 09:04 PM, Andrew Cooper wrote:
On 08/06/15 04:45, Yang Hongyang wrote:
call callbacks->get_dirty_pfn() after suspend primary vm to
get dirty pages on secondary vm, and send pages both dirty on
primary/secondary to secondary.
Signed-off-by: Yang Hongyang
Signed-off-by: Wen Congy
flight 58175 linux-next real [real]
http://logs.test-lab.xenproject.org/osstest/logs/58175/
Regressions :-(
Tests which did not succeed and are blocking,
including tests which could not be run:
test-amd64-amd64-rumpuserxen-amd64 3 host-install(3) broken REGR. vs. 57931
test-amd64-amd64-xl-cr
flight 58185 linux-3.18 real [real]
http://logs.test-lab.xenproject.org/osstest/logs/58185/
Regressions :-(
Tests which did not succeed and are blocking,
including tests which could not be run:
test-amd64-i386-xl-qemuu-win7-amd64 16 guest-stop fail in 58064 REGR. vs. 57312
Tests which are faili
On 06/08/2015 06:15 PM, Andrew Cooper wrote:
On 08/06/15 10:58, Yang Hongyang wrote:
On 06/08/2015 05:46 PM, Andrew Cooper wrote:
On 08/06/15 04:43, Yang Hongyang wrote:
ioreq page contains evtchn which will be set when we resume the
secondary vm the first time. The hypervisor will check i
From: "Luis R. Rodriguez"
We are burrying direct access to MTRR code support on
x86 in order to take advantage of PAT. In the future we
also want to make the default behaviour of ioremap_nocache()
to use strong UC, use of mtrr_add() on those systems
would make write-combining void.
In order to h
flight 58167 linux-linus real [real]
http://logs.test-lab.xenproject.org/osstest/logs/58167/
Failures and problems with tests :-(
Tests which did not succeed and are blocking,
including tests which could not be run:
test-amd64-amd64-rumpuserxen-amd64 3 host-install(3) broken REGR. vs. 58128
From: "Luis R. Rodriguez"
We are burrying direct access to MTRR code support on
x86 in order to take advantage of PAT. In the future we
also want to make the default behaviour of ioremap_nocache()
to use strong UC, use of mtrr_add() on those systems
would make write-combining void.
In order to h
On Thu, May 28, 2015 at 2:50 PM, Luis R. Rodriguez
wrote:
> On Thu, May 21, 2015 at 11:32 AM, Luis R. Rodriguez wrote:
>> On Thu, May 21, 2015 at 04:20:27PM +0800, Michal Marek wrote:
>>> Dne 21.5.2015 v 02:53 Luis R. Rodriguez napsal(a):
>>> > From: "Luis R. Rodriguez"
>>> >
>>> > Michal Marek,
Hi Parth,
On 17/05/2015 21:04, Parth Dixit wrote:
ACPI will be disabled by default. Define new
command line parameter "acpi" for enabling it.
I don't think this is right. We have to be able to boot platform where
there is only ACPI provided without adding a command line option.
Overall, it
Hi,
On 17/05/2015 21:04, Parth Dixit wrote:
NOTE: This is a wrokaround to be fixed later.
How do you plan to fix it?
Route all the irq's to Dom0 at the time of booting.
Trigger and polarity will be set dyanmaically when
s/dyanmaically/dynamically/
Dom0 request's for it.
Signed-off-by: P
On 17/05/2015 21:04, Parth Dixit wrote:
Map acpi tables described in uefi table to DOM0 address space
Signed-off-by: Parth Dixit
---
xen/arch/arm/domain_build.c | 59 -
1 file changed, 58 insertions(+), 1 deletion(-)
diff --git a/xen/arch/arm/do
Hi Parth,
On 17/05/2015 21:03, Parth Dixit wrote:
creates a helper function for mapping with cached attributes
Signed-off-by: Parth Dixit
---
xen/arch/arm/p2m.c| 26 ++
xen/include/asm-arm/p2m.h | 10 ++
2 files changed, 36 insertions(+)
diff --git
Hi Parth,
On 17/05/2015 21:03, Parth Dixit wrote:
EFI table, memory description table and some of acpi tables
will reside in DOM0 memory. Add placeholder for starting
address for loading in DOM0 and get/set acpi size helpers.
Signed-off-by: Parth Dixit
---
xen/arch/arm/acpi/lib.c| 12 +++
Hi Parth,
On 17/05/2015 21:03, Parth Dixit wrote:
In ACPI mmio regions are described in DSDT which requires
AML interpreter. Defer the mapping of mmio until DSDT is parsed in
DOM0 and map mmio's dynamically at the time of request.
I'm against a such solution. Even though it's DOM0 we should av
Hi Parth,
I think it misses a ":" between acpi and estimate.
On 17/05/2015 21:03, Parth Dixit wrote:
Estimate the memory required for loading acpi/efi tablee
in DOM0. Initialize the size of acpi/efi tables.
It needs more description...
Signed-off-by: Parth Dixit
---
xen/arch/arm/domain_
Hi,
On 17/05/2015 21:04, Parth Dixit wrote:
Add helper functions for calculating crc32.
This is required for computing crc of efi table.
These functions are copied from here
http://mirrors.neusoft.edu.cn/rpi-kernel/lib/xz/xz_crc32.c
Original author's are Lasse Collin and Igor Pavlov.
I'm near
Hi Parth,
On 17/05/2015 21:03, Parth Dixit wrote:
ACPI memory is seperate from conventional memory and should
s/seperate/separate/
be marked as reserved while passing to DOM0. Create a new meminfo
structure to store all the acpi tables listed in uefi.
Signed-off-by: Parth Dixit
---
xen/a
Hi Parth,
On 17/05/2015 21:04, Parth Dixit wrote:
Interrupt information is described in DSDT and is not available at
the time of booting. Configure the interrupts dynamically when requested
by Dom0
Missing "."
Also, I'm sure we talked about it multiple time. I'd like to keep the
ACPI changes
flight 58161 libvirt real [real]
http://logs.test-lab.xenproject.org/osstest/logs/58161/
Regressions :-(
Tests which did not succeed and are blocking,
including tests which could not be run:
test-amd64-i386-libvirt 11 guest-start fail REGR. vs. 53854
test-amd64-amd64-libvirt
changes v1 to v2:
Split v1 patch into 3.
Added a BUG FIX patch (#1: "exec: Do not use MemoryRegion after
free").
Technically this bug fix should be a separate patch, however this
issue only seems to reproduce when this patch set is applied.
Michael S. Tsirkin:
"You need some
Here is gdb output that shows this happening:
Breakpoint 3, object_finalize (data=0x7fdf32a14010) at qom/object.c:417
417 obj->free(obj);
(gdb) bt
#0 object_finalize (data=0x7fdf32a14010) at qom/object.c:417
#1 0x007329d4 in object_unref (obj=0x7fdf32a14010) at
qom
The commit 707ff80021ccd7a68f4b3d2c44eebf87efbb41c4 added usage of
xc_hvm_map_pcidev_from_ioreq_server() and
xc_hvm_unmap_pcidev_from_ioreq_server(). These routines only
correctly work if the PCI BDF of a PCI device is unique. The 3
parts (bus, device, and function) of a PCI BDF are not required
Signed-off-by: Don Slutz
CC: Don Slutz
---
include/hw/xen/xen_common.h | 10 ++
xen-hvm.c | 13 -
2 files changed, 18 insertions(+), 5 deletions(-)
diff --git a/include/hw/xen/xen_common.h b/include/hw/xen/xen_common.h
index 38f29fb..6579b78 100644
--- a/in
The commit 707ff80021ccd7a68f4b3d2c44eebf87efbb41c4 assumed that a
PCI device has a static address. This is not true for PCI devices
that are on the secondary bus of a PCI to PCI bridge.
BIOS and/or guest OS can change the secondary bus number to a new
value at any time.
Extend the device listen
On Mon, Jun 8, 2015 at 11:21 AM, Dario Faggioli
wrote:
> On Mon, 2015-05-25 at 19:11 -0500, Chong Li wrote:
>> Change main_sched_rtds and related output functions to support per-VCPU
>> settings
>> for xl sched-rtds tool.
>>
>> Signed-off-by: Chong Li
>> Signed-off-by: Meng Xu
>> Signed-off-by:
On Mon, 2015-06-08 at 14:01 +0100, Lars Kurth wrote:
> > On 8 Jun 2015, at 13:19, Ian Campbell wrote:
> >
> > In MAINTAINERS S: Supported means:
> >
> > "Someone is actually paid to look after this.", which I think is
> > distinct from "This works well enough that the project is happy to
> > rec
On 06/07/15 14:29, Wei Liu wrote:
> On Mon, Jun 01, 2015 at 09:13:08AM +, Maoming wrote:
>> Hi all:
>>I encountered a troublesome problem about OVMF.
>>I used OVMF.fd as a BIOS of virtual machine(VM).
>>
>>1、my environment:
>>xen_version: 4.6-unstable
On Mon, Jun 8, 2015 at 10:56 AM, Dario Faggioli
wrote:
> On Fri, 2015-06-05 at 12:37 +0100, Ian Campbell wrote:
>> On Mon, 2015-05-25 at 19:09 -0500, Chong Li wrote:
>
>> > diff --git a/tools/libxl/libxl_types.idl b/tools/libxl/libxl_types.idl
>> > index 117b61d..d28d274 100644
>> > --- a/tools/li
Hi Julien,
>My test was limited as I don't have a platform where CNTFRQ/CNTFRQ_EL0 is not
>valid. I may have done a mistake in the code.
Understood. That's why I thought it would be worthwhile posting my results :-)
>> What I see is that in preinit_xen_time(), the call to dt_property_read_u32()
Currently only QEMU traditional supports stubdom, so we only create
test-amd64-amd64-xl-qemut-stubdom-debianhvm-amd64
test-amd64-amd64-xl-qemut-stubdom-debianhvm-amd64-xsm
test-amd64-i386-xl-qemut-stubdom-debianhvm-amd64
test-amd64-i386-xl-qemut-stubdom-debianhvm-amd64-xsm
Note that stubdom only
...
> == Hypervisor ==
[...]
>
> * Improve RTDS scheduler (none)
>Change RTDS from quantum driven to event driven
> - Dagaen Golomb, Meng Xu, Chong Li
>
...
Ok.
The patch for this is out:
http://osdir.com/ml/general/2015-06/msg10265.html
Looking forward to comments.
Regards,
Dagaen Go
On Mon, 8 Jun 2015, Ian Jackson wrote:
> Stefano Stabellini writes ("Re: [PATCH v2 5/5] libxl: spawns two QEMUs for
> HVM guests"):
> > Do you have any suggestions on how I could fix this function for the
> > purpose of this series? I would be quite happy to leave fixing
> > qdisk_spawn_outcome in
On Mon, Jun 08, 2015 at 04:53:46PM +0100, Wei Liu wrote:
> On Wed, Jun 03, 2015 at 03:35:18PM +0200, Vitaly Kuznetsov wrote:
> > Jan and Tim,
> >
> > last week you expressed some concerns about if the toolstack-based
> > approach to PVHVM guest kexec is the best. Here you can see the 'reset
> > ev
flight 58156 ovmf real [real]
http://logs.test-lab.xenproject.org/osstest/logs/58156/
Regressions :-(
Tests which did not succeed and are blocking,
including tests which could not be run:
test-amd64-amd64-xl-qemuu-win7-amd64 16 guest-stopfail REGR. vs. 56492
test-amd64-i386-xl-qemuu-win
flight 58151 xen-unstable real [real]
http://logs.test-lab.xenproject.org/osstest/logs/58151/
Failures :-/ but no regressions.
Tests which are failing intermittently (not blocking):
test-amd64-i386-libvirt 6 xen-bootfail pass in 58095
test-armhf-armhf-xl 9 de
Signed-off-by: Anthony PERARD
---
src/libxl/libxl_conf.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/libxl/libxl_conf.c b/src/libxl/libxl_conf.c
index 9b258ac..e845759 100644
--- a/src/libxl/libxl_conf.c
+++ b/src/libxl/libxl_conf.c
@@ -1541,7 +1541,7 @@ libxlDriverCon
Hello,
On 6/8/2015 9:24 AM, Jan Beulich wrote:
> arinc653: don't leak hypervisor stack contents through
> XEN_SYSCTL_SCHEDOP_getinfo
>
> Note that due to XSA-77 this is not a security issue.
>
> Reported-by: "栾尚聪(好风)"
> Signed-off-by: Jan Beulich
>
> --- a/xen/common/sched_arinc653.c
> ++
Stefano Stabellini writes ("Re: [PATCH v2 5/5] libxl: spawns two QEMUs for HVM
guests"):
> Do you have any suggestions on how I could fix this function for the
> purpose of this series? I would be quite happy to leave fixing
> qdisk_spawn_outcome in libxl.c to you :-)
Sure.
> TBH as I wrote earl
On Mon, 8 Jun 2015, Ian Jackson wrote:
> Stefano Stabellini writes ("[PATCH v2 3/5] libxl: xsrestrict QEMU"):
> > Check whether QEMU supports the xsrestrict option, by parsing its --help
> > output. Store the result on xenstore for future reference on a per QEMU
> > binary basis, so that device_mod
On Mon, 2015-05-25 at 19:11 -0500, Chong Li wrote:
> Change main_sched_rtds and related output functions to support per-VCPU
> settings
> for xl sched-rtds tool.
>
> Signed-off-by: Chong Li
> Signed-off-by: Meng Xu
> Signed-off-by: Sisu Xi
> ---
> tools/libxl/xl_cmdimpl.c | 261
> +++
On Mon, 8 Jun 2015, Ian Jackson wrote:
> Stefano Stabellini writes ("Re: [PATCH v2 5/5] libxl: spawns two QEMUs for
> HVM guests"):
> > On Sun, 7 Jun 2015, Wei Liu wrote:
> > > Also it looks very similar to the function with the same name in
> > > libxl.c. I'm not sure why that doesn't propagate r
On 06/08/15 11:37, George Dunlap wrote:
> On 06/08/2015 04:01 PM, Don Slutz wrote:
>> On 06/08/15 10:20, George Dunlap wrote:
>>> And at the moment, pages in the p2m are allocated by a number of entities:
>>> * In the libxc domain builder.
>>> * In the guest balloon driver
>>> * And now, in qemu, t
"Jan Beulich" writes:
On 03.06.15 at 15:35, wrote:
>> +int arch_domain_soft_reset(struct domain *d)
>> +{
>> +struct page_info *page = virt_to_page(d->shared_info), *new_page;
>> +int ret = 0;
>> +struct domain *owner;
>> +unsigned long mfn, mfn_new, gfn;
>> +p2m_type_t
On Fri, 2015-06-05 at 12:37 +0100, Ian Campbell wrote:
> On Mon, 2015-05-25 at 19:09 -0500, Chong Li wrote:
> > diff --git a/tools/libxl/libxl_types.idl b/tools/libxl/libxl_types.idl
> > index 117b61d..d28d274 100644
> > --- a/tools/libxl/libxl_types.idl
> > +++ b/tools/libxl/libxl_types.idl
> > @
On Wed, Jun 03, 2015 at 03:35:18PM +0200, Vitaly Kuznetsov wrote:
> Jan and Tim,
>
> last week you expressed some concerns about if the toolstack-based
> approach to PVHVM guest kexec is the best. Here you can see the 'reset
> everything' approach to the same problem. It is the bare minimum of wha
Stefano Stabellini writes ("[PATCH v2 3/5] libxl: xsrestrict QEMU"):
> Check whether QEMU supports the xsrestrict option, by parsing its --help
> output. Store the result on xenstore for future reference on a per QEMU
> binary basis, so that device_model_override still works fine with it.
...
> +
Stefano Stabellini writes ("[PATCH v2 1/5] libxl: allow
/local/domain/0/device-model/$DOMID to be written by $DOMID"):
> The device model is going to restrict its xenstore connection to $DOMID
> level. Let it access /local/domain/0/device-model/$DOMID, as it is
> required by QEMU to read/write the
Stefano Stabellini writes ("Re: [PATCH v2 5/5] libxl: spawns two QEMUs for HVM
guests"):
> On Sun, 7 Jun 2015, Wei Liu wrote:
> > Also it looks very similar to the function with the same name in
> > libxl.c. I'm not sure why that doesn't propagate rc either...
I think that's a bug.
> Right, I ba
On Mon, Jun 08, 2015 at 04:20:48PM +0100, George Dunlap wrote:
> On 06/08/2015 03:53 PM, Konrad Rzeszutek Wilk wrote:
> > On Mon, Jun 08, 2015 at 01:11:38PM +0100, Ian Campbell wrote:
> >> On Mon, 2015-06-08 at 12:40 +0100, Stefano Stabellini wrote:
> >>
> > I disagree that libxl should be the
flight 58150 qemu-mainline real [real]
http://logs.test-lab.xenproject.org/osstest/logs/58150/
Regressions :-(
Tests which did not succeed and are blocking,
including tests which could not be run:
test-amd64-amd64-xl-qemuu-debianhvm-amd64-xsm 11 guest-saverestore fail REGR.
vs. 57815
test-amd6
Vitaly Kuznetsov writes ("[PATCH RFC 0/4] 'reset everything' approach to PVHVM
guest kexec"):
> 1) As XS_RESET_WATCHES is not supported by oxenstored we need to try removing
> the watch in case add operation failed, e.g.:
It would surely be better to fix oxenstored. IIRC the approach of
trying t
On 06/08/2015 04:01 PM, Don Slutz wrote:
> On 06/08/15 10:20, George Dunlap wrote:
>> And at the moment, pages in the p2m are allocated by a number of entities:
>> * In the libxc domain builder.
>> * In the guest balloon driver
>> * And now, in qemu, to allocate extra memory for virtual ROMs.
>
>
>>> On 08.06.15 at 16:58, wrote:
> "Jan Beulich" writes:
>
> On 03.06.15 at 15:35, wrote:
>>> When soft reset is being performed we need to replace all actively
>>> granted pages with empty pages to prevent possible future memory
>>> corruption as the newly started kernel won't be aware of
>>> On 03.06.15 at 15:35, wrote:
> +int arch_domain_soft_reset(struct domain *d)
> +{
> +struct page_info *page = virt_to_page(d->shared_info), *new_page;
> +int ret = 0;
> +struct domain *owner;
> +unsigned long mfn, mfn_new, gfn;
> +p2m_type_t __p2mt;
No need for leading und
On Mon, 2015-06-08 at 11:21 -0400, Konrad Rzeszutek Wilk wrote:
> On Mon, Jun 08, 2015 at 03:47:22PM +0100, Ian Jackson wrote:
> > Konrad Rzeszutek Wilk writes ("Re: [Xen-devel] [xen-unstable test] 57852:
> > regressions - FAIL"):
> > > Could it be an missing microcode update? I don't know if the
On Mon, 8 Jun 2015, Wei Liu wrote:
> On Thu, Jun 04, 2015 at 12:28:18PM +0100, Stefano Stabellini wrote:
> > If QEMU is ran just to provide PV backends, change the xenstore path to
> > /local/domain/0/device-model/$DOMID/pv.
> >
> > Add a parameter to libxl__device_model_xs_path to distinguish the
On Mon, Jun 08, 2015 at 03:47:22PM +0100, Ian Jackson wrote:
> Konrad Rzeszutek Wilk writes ("Re: [Xen-devel] [xen-unstable test] 57852:
> regressions - FAIL"):
> > Could it be an missing microcode update? I don't know if the OSSTest does
> > the ucode=scan or updates the microcode later?
>
> I t
On 06/08/2015 03:53 PM, Konrad Rzeszutek Wilk wrote:
> On Mon, Jun 08, 2015 at 01:11:38PM +0100, Ian Campbell wrote:
>> On Mon, 2015-06-08 at 12:40 +0100, Stefano Stabellini wrote:
>>
> I disagree that libxl should be the arbitrator of a property that is
> stored, maintained and enforced by
. snip ..
> >> * You may need support for adapting or augmenting exception tables if
> >> patching such code is desired (it probably is). Hotpatches may need
> >> to bring their own small exception tables (similar to how Linux
> >> modules support this). If you don't plan on supporting hotp
On Sun, 7 Jun 2015, Wei Liu wrote:
> On Thu, Jun 04, 2015 at 12:28:19PM +0100, Stefano Stabellini wrote:
> > Starts a second QEMU to provide PV backends in userspace to HVM guests.
> >
> > Signed-off-by: Stefano Stabellini
> > ---
> > tools/libxl/libxl_create.c | 18 ++
> > too
On Mon, Jun 08, 2015 at 04:44:05PM +0200, Fabio Fantoni wrote:
[...]
> >>config/Paths.mk instead contain it correctly.
> >>I checked the commit
> >>http://xenbits.xen.org/gitweb/?p=xen.git;a=commitdiff;h=bf36d0792032a0245e4a17a64c8b9eb9e056f282
> >>but not add something about libxl/_paths.h and in
"Jan Beulich" writes:
On 03.06.15 at 15:35, wrote:
>> --- a/xen/common/event_channel.c
>> +++ b/xen/common/event_channel.c
>> @@ -274,11 +274,13 @@ static long
>> evtchn_bind_interdomain(evtchn_bind_interdomain_t *bind)
>>
>> lchn->u.interdomain.remote_dom = rd;
>> lchn->u.int
"Jan Beulich" writes:
On 03.06.15 at 15:35, wrote:
>> When soft reset is being performed we need to replace all actively
>> granted pages with empty pages to prevent possible future memory
>> corruption as the newly started kernel won't be aware of these
>> granted pages.
>>
>> We make the
On 06/08/15 10:20, George Dunlap wrote:
> On 06/08/2015 02:22 PM, Stefano Stabellini wrote:
>>> 3. A group of entities which operate in isolation by only ever
>>> increasing or descreasing the max pages according to their own
>>> requirements, without reference to anyone else.
If hvmemul_do_io_addr() is called to complete a previously issued
emulation then there is no need to acquire the RAM pages again. There
is also no need to re-calculate the value of *reps, providing
hvmemul_do_io() updates it when returning X86EMUL_OKAY.
Signed-off-by: Paul Durrant
Cc: Keir Fraser
The code in hvmemul_do_io() that tracks large reads or writes, to avoid
re-issue of component I/O, is defeated by accesses across a page boundary
because it uses physical rather than linear address. The code is also only
relevant to memory mapped I/O to or from a buffer.
This patch re-factors the
Konrad Rzeszutek Wilk writes ("Re: [Xen-devel] [xen-unstable test] 57852:
regressions - FAIL"):
> Could it be an missing microcode update? I don't know if the OSSTest does
> the ucode=scan or updates the microcode later?
I think osstest's machines don't get microcode updates. I'm no expert
on x8
Jan Beulich writes ("Re: [Xen-devel] [xen-unstable test] 57852: regressions -
FAIL"):
> On 08.06.15 at 11:27, wrote:
> > I don't know much about the hardware in the pool other than what can be
> > gathered from the serial and dmesg logs.
>
> Right - this is useful for learning details of an indi
...in hvmemul_read/write()
Add hvmemul_phys_mmio_access() and hvmemul_linear_mmio_access() functions
to reduce code duplication.
Signed-off-by: Paul Durrant
Cc: Keir Fraser
Cc: Jan Beulich
Cc: Andrew Cooper
---
xen/arch/x86/hvm/emulate.c | 237 +---
1
Emulation request status is already covered by STATE_IOREQ_XXX values so
just use those. The mapping is:
HVMIO_none-> STATE_IOREQ_NONE
HVMIO_awaiting_completion -> STATE_IOREQ_READY
HVMIO_completed -> STATE_IORESP_READY
Signed-off-by: Paul Durrant
Cc: Keir Fraser
Cc: J
...otherwise they will simply carry on to the next page using a normal
linear-to-phys translation.
Signed-off-by: Paul Durrant
Cc: Keir Fraser
Cc: Jan Beulich
Cc: Andrew Cooper
---
xen/arch/x86/hvm/emulate.c |6 --
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/xen/arch
The is_mmio parameter can be inferred from the ioreq type.
Signed-off-by: Paul Durrant
Cc: Keir Fraser
Cc: Jan Beulich
Cc: Andrew Cooper
---
xen/arch/x86/hvm/emulate.c |7 ---
1 file changed, 4 insertions(+), 3 deletions(-)
diff --git a/xen/arch/x86/hvm/emulate.c b/xen/arch/x86/hvm/e
Use an ioreq_t rather than open coded state, size, dir and data fields
in struct hvm_vcpu_io. This also allows PIO completion to be handled
similarly to MMIO completion by re-issuing the handle_pio() call.
Signed-off-by: Paul Durrant
Cc: Keir Fraser
Cc: Jan Beulich
Cc: Andrew Cooper
---
xen/a
By removing the HVMIO_dispatched state and making all pending emulations
(i.e. all those not handled by the hypervisor) use HVMIO_awating_completion,
various code-paths can be simplified.
Signed-off-by: Paul Durrant
Cc: Keir Fraser
Cc: Jan Beulich
Cc: Andrew Cooper
---
xen/arch/x86/hvm/emulat
On Mon, Jun 08, 2015 at 01:11:38PM +0100, Ian Campbell wrote:
> On Mon, 2015-06-08 at 12:40 +0100, Stefano Stabellini wrote:
>
> > > > I disagree that libxl should be the arbitrator of a property that is
> > > > stored, maintained and enforced by Xen. Xen should be the arbitrator.
> > >
> > > Tha
On Sun, 7 Jun 2015, Wei Liu wrote:
> On Thu, Jun 04, 2015 at 12:28:15PM +0100, Stefano Stabellini wrote:
> > The device model is going to restrict its xenstore connection to $DOMID
> > level. Let it access /local/domain/0/device-model/$DOMID, as it is
> > required by QEMU to read/write the physmap.
Il 08/06/2015 15:56, Wei Liu ha scritto:
On Mon, Jun 08, 2015 at 03:49:24PM +0200, Fabio Fantoni wrote:
Il 08/06/2015 15:30, Wei Liu ha scritto:
On Mon, Jun 08, 2015 at 02:36:09PM +0200, Fabio Fantoni wrote:
Il 08/06/2015 13:28, Wei Liu ha scritto:
On Mon, Jun 08, 2015 at 01:25:05PM +0200, Fa
Hi,
some inline comments.
On 05.06.2015 17:00, Konrad Rzeszutek Wilk wrote:
> On Wed, May 20, 2015 at 05:11:20PM +0200, Martin Pohlack wrote:
>> Hi,
>>
>> this looks very interesting.
>
> Thank you!
>>
>> I have talked about an experimental Xen hotpatching design at Linux
>> Plumbers Conference
The implementation of mmio and portio intercepts is unnecessarily different.
This leads to much code duplication. This patch unifies much of the
intercept handling, leaving only distinct handlers for stdvga mmio and dpci
portio. Subsequent patches will unify those handlers.
Signed-off-by: Paul Dur
>>> On 08.06.15 at 16:29, wrote:
> On Mon, 2015-06-08 at 14:28 +0100, Jan Beulich wrote:
>> Being a pre-3.1 compatibility hypercall handler only, it's not needed
>> on ARM or any future architectures Xen may get ported to.
>>
>> Also the function shouldn't really be used internally - its use shou
The check is done at the wrong point (since it is irrelevant if the
I/O is to be handled by the hypervisor) and its functionality can be
covered by returning X86EMUL_UNHANDLEABLE from hvm_send_assist_req()
instead.
This patch also removes the domain_crash() call from
hvm_send_assist_req(). Returni
This patch re-works the dpci portio intercepts so that they can be unified
with standard portio handling thereby removing a substantial amount of
code duplication.
Signed-off-by: Paul Durrant
Cc: Keir Fraser
Cc: Jan Beulich
Cc: Andrew Cooper
---
xen/arch/x86/hvm/hvm.c |2 +
xen/
By removing the calls in hvmemul_do_io() (which is replaced by a single
assignment) and hvm_complete_assist_request() (which is replaced by a
call to process_portio_intercept() with a suitable set of ops) then
hvm_io_assist() can be moved into hvm.c and made static (and hence be a
candidate for inl
This patch series re-works much of the code involved in emulation of port
and memory mapped I/O for HVM guests.
The code has become very convoluted and, at least by inspection, certain
emulations will apparently malfunction.
The series is broken down into 17 patches (which are also available in
m
Currently hvmemul_do_io() handles paging for I/O to/from a guest address
inline. This causes every exit point to have to execute:
if ( ram_page )
put_page(ram_page);
This patch introduces wrapper hvmemul_do_io_addr() and
hvmemul_do_io_buffer() functions. The latter is used for I/O to/from a X
...and error handling"
NOTE: A straight reversion was not possible because of subsequent changes
in the code so this at least partially a manual reversion.
By limiting hvmemul_do_io_addr() to reps falling within the pages on which
a reference has already been taken, we can guarantee that ca
The struct just contains three methods and no data, so the name
hvm_mmio_ops more accurately reflects its content. A subsequent patch
introduces a new structure which more accurately warrants the name
hvm_mmio_handler so doing the rename in this purely cosmetic patch avoids
conflating functional an
It's clear from the following check in hvmemul_rep_movs:
if ( sp2mt == p2m_mmio_direct || dp2mt == p2m_mmio_direct ||
(sp2mt == p2m_mmio_dm && dp2mt == p2m_mmio_dm) )
return X86EMUL_UNHANDLEABLE;
that mmio <-> mmio copy is not handled. This means the code in the
stdvga mmio i
The state of in-flight I/O and how its completion will be handled are
logically separate and conflating the two makes the code unnecessarily
confusing.
Signed-off-by: Paul Durrant
Cc: Keir Fraser
Cc: Jan Beulich
Cc: Andrew Cooper
---
xen/arch/x86/hvm/hvm.c | 42 +
>>> On 03.06.15 at 15:35, wrote:
> --- a/xen/common/domain.c
> +++ b/xen/common/domain.c
> @@ -1116,12 +1116,12 @@ int map_vcpu_info(struct vcpu *v, unsigned long gfn,
> unsigned offset)
>
> /*
> * Unmap the vcpu info page if the guest decided to place it somewhere
> - * else. This is only
On Mon, 2015-06-08 at 14:28 +0100, Jan Beulich wrote:
> Being a pre-3.1 compatibility hypercall handler only, it's not needed
> on ARM or any future architectures Xen may get ported to.
>
> Also the function shouldn't really be used internally - its use should
> be limited to its purpose (and henc
On Sat, Jun 06, 2015 at 11:57:31AM +0200, Roger Pau Monné wrote:
> El 05/06/15 a les 23.52, Tim Deegan ha escrit:
> > At 18:21 +0100 on 05 Jun (1433528517), Andrew Cooper wrote:
> >> On 05/06/15 18:16, Stefano Stabellini wrote:
> >>> On Fri, 5 Jun 2015, Andrew Cooper wrote:
> On 05/06/15 17:43
1 - 100 of 250 matches
Mail list logo
