>>> On 11.02.16 at 22:10, wrote:
> c/s 8135cf8b092723dbfcc611fe6fdcb3a36c9951c5
> "xen/pciback: Save xen_pci_op commands before processing it"
> would copyback the processed values - which was great.
>
> However it missed the case that xen_pcibk_enable_msix - when
>
Moving a vCPU to a different pCPU means offlining it and
then waking it up, on the new pCPU. Credit1 grants BOOST
priority to vCPUs that wakes up, with the aim of improving
I/O latency. The net effect of this all is that vCPUs get
boosted when migrating, which shouldn't happen.
For instance, this
>>> On 11.02.16 at 20:25, wrote:
> CentOS 7 gets into trouble when compiling Xen citing:
>
> flushtlb.c: Assembler messages:
> flushtlb.c:149: Error: value of 256 too large for field of 1 bytes at 1
>
> The line number is wrong, and the error message not helpful.
>>> On 12.02.16 at 10:04, wrote:
> --- a/xen/include/public/memory.h
> +++ b/xen/include/public/memory.h
> @@ -390,8 +390,6 @@ DEFINE_XEN_GUEST_HANDLE(xen_mem_paging_op_t);
> #define XENMEM_access_op21
> #define XENMEM_access_op_set_access
Hi again,
Here it comes v2, redone following Jan's suggestion, which allowed to get rid
of patch 2, and do everything in sched_credit.c.
So, in summary, because of the fact that vcpu_migrate() forces the vcpus into a
sleep+wakeup cycle, vcpus being migrated to a new pcpu, were also being granted
I noticed that the qemu(1) links in
http://xenbits.xen.org/docs/unstable/man/xl.cfg.5.html (and probably other
docs) are broken. They go to http://man.he.net/man1/qemu, which does not exist
any more
Lars
___
Xen-devel mailing list
Add tracepoints and a performance counter for
boosting and unboosting in Credit1.
Note that they (the trace points) do not cover
the case of the idle vCPU being boosted to run
a tasklet, as there already is
TRC_CSCHED_SCHED_TASKLET for that.
Signed-off-by: Dario Faggioli
On 12/02/16 08:13, Jan Beulich wrote:
On 11.02.16 at 20:41, wrote:
>> On 11/02/16 19:25, Andrew Cooper wrote:
>>> --- a/xen/arch/x86/flushtlb.c
>>> +++ b/xen/arch/x86/flushtlb.c
>>> @@ -141,10 +141,10 @@ void flush_area_local(const void *va, unsigned int
>>>
>>> On 12.02.16 at 10:51, wrote:
> On 12/02/16 08:23, Jan Beulich wrote:
> On 11.02.16 at 20:25, wrote:
>>> CentOS 7 gets into trouble when compiling Xen citing:
>>>
>>> flushtlb.c: Assembler messages:
>>> flushtlb.c:149: Error: value
On 02/12/2016 11:10 AM, Jan Beulich wrote:
On 12.02.16 at 10:04, wrote:
>> --- a/xen/include/public/memory.h
>> +++ b/xen/include/public/memory.h
>> @@ -390,8 +390,6 @@ DEFINE_XEN_GUEST_HANDLE(xen_mem_paging_op_t);
>> #define XENMEM_access_op21
>>> On 11.02.16 at 20:51, wrote:
> Otherwise, debug code such as "void __attribute__((noreturn)) foobar()" fails
> to compile when the noreturn itself gets expanded, resulting in
> __attribute__((__attribute__((noreturn.
Well, why would the debugging code not use
>>> On 12.02.16 at 10:37, wrote:
> @@ -787,6 +788,16 @@ _csched_cpu_pick(const struct scheduler *ops, struct
> vcpu *vc, bool_t commit)
> static int
> csched_cpu_pick(const struct scheduler *ops, struct vcpu *vc)
> {
> +struct csched_vcpu *svc = CSCHED_VCPU(vc);
On 12/02/16 08:23, Jan Beulich wrote:
On 11.02.16 at 20:25, wrote:
>> CentOS 7 gets into trouble when compiling Xen citing:
>>
>> flushtlb.c: Assembler messages:
>> flushtlb.c:149: Error: value of 256 too large for field of 1 bytes at 1
>>
>> The line number is
Remove the "prepare for reconnect" pr_info in xenbus.c. It's largely
uninteresting and the states of the frontend and backend can easily be
observed by watching the (o)xenstored log.
Signed-off-by: Paul Durrant
Cc: Ian Campbell
Cc: Wei Liu
The code does not currently support a frontend passing multiple extra info
fragments to the backend in a tx request. The xenvif_get_extras() function
handles multiple extra_info fragments but make_tx_response() assumes there
is only ever a single extra info fragment.
This patch modifies
On 02/12/2016 11:57 AM, Jan Beulich wrote:
On 12.02.16 at 01:22, wrote:
>> --- a/xen/arch/x86/vm_event.c
>> +++ b/xen/arch/x86/vm_event.c
>> @@ -122,6 +122,65 @@ void vm_event_set_registers(struct vcpu *v,
>> vm_event_response_t *rsp)
>> v->arch.user_regs.eip =
>>> On 12.02.16 at 04:08, wrote:
> in the keyhandler.h file. Otherwise on ARM builds if we
> just use the keyhandler file - the compile will fail.
>
> CC: ian.campb...@citrix.com
> CC: wei.l...@citrix.com
> CC: stefano.stabell...@citrix.com
> Signed-off-by: Konrad
>>> On 12.02.16 at 01:22, wrote:
> Sending the dr7 register during vm_events is useful for various applications,
> but the current way the register value is gathered is incorrent. In this
> patch
> we extend vmx_vmcs_save so that we get the correct value.
>
> Suggested-by:
>>> On 11.02.16 at 20:41, wrote:
> On 11/02/16 19:25, Andrew Cooper wrote:
>> --- a/xen/arch/x86/flushtlb.c
>> +++ b/xen/arch/x86/flushtlb.c
>> @@ -141,10 +141,10 @@ void flush_area_local(const void *va, unsigned int
>> flags)
>> {
>>
On 2/12/2016 8:05 AM, Corneliu ZUZU wrote:
On 2/11/2016 5:44 PM, Tamas K Lengyel wrote:
* the #ifdefs make it possible for that code to be put in common
=> that makes it *clear* that those code parts are NOT
architecture specific and their implementation can be safely used
for
Hey Juergen,
On Fri, Feb 12, 2016 at 07:25:02AM +0100, Juergen Gross wrote:
[...]
> Okay, let me do some cleanup work on the xen loader:
>
> - add the possibility to call it multiple times (state reset, free the
> allocated memory)
> - merge all necessary global variables into one state
xc_mem_access_enable_emulate() and xc_mem_access_disable_emulate()
are currently no-ops, that is all they do is set a flag that
nobody else checks. The user can already set the EMULATE flags in
the vm_event response if emulation is desired, and having an extra
check above that is not inherently
>>> On 11.02.16 at 22:10, wrote:
> c/s 408fb0e5aa7fda0059db282ff58c3b2a4278baa0
> "xen/pciback: Don't allow MSI-X ops if PCI_COMMAND_MEMORY is not set."
> would check the device for PCI_COMMAND_MEMORY which is great.
> Except that VF devices are unique - for example they
On 12/02/16 10:00, Jan Beulich wrote:
On 12.02.16 at 10:51, wrote:
>> On 12/02/16 08:23, Jan Beulich wrote:
>> On 11.02.16 at 20:25, wrote:
CentOS 7 gets into trouble when compiling Xen citing:
flushtlb.c: Assembler
>>> On 12.02.16 at 11:02, wrote:
> On 12/02/16 10:00, Jan Beulich wrote:
> On 12.02.16 at 10:51, wrote:
>>> On 12/02/16 08:23, Jan Beulich wrote:
>>> On 11.02.16 at 20:25, wrote:
> CentOS 7 gets into
>>> On 12.02.16 at 11:19, wrote:
> On 02/12/2016 11:57 AM, Jan Beulich wrote:
> On 12.02.16 at 01:22, wrote:
>>> --- a/xen/arch/x86/vm_event.c
>>> +++ b/xen/arch/x86/vm_event.c
>>> @@ -122,6 +122,65 @@ void vm_event_set_registers(struct vcpu
On Fri, 2016-02-12 at 02:50 -0700, Jan Beulich wrote:
> > > > On 12.02.16 at 10:37, wrote:
> > @@ -787,6 +788,16 @@ _csched_cpu_pick(const struct scheduler *ops,
> > struct vcpu *vc, bool_t commit)
> > static int
> > csched_cpu_pick(const struct scheduler *ops, struct
On 12/02/16 10:12, Jan Beulich wrote:
On 12.02.16 at 11:02, wrote:
>> On 12/02/16 10:00, Jan Beulich wrote:
>> On 12.02.16 at 10:51, wrote:
On 12/02/16 08:23, Jan Beulich wrote:
On 11.02.16 at 20:25,
On Fri, 2016-02-12 at 16:22 +0530, Harmandeep Kaur wrote:
> On Thu, Feb 11, 2016 at 3:18 PM, Dario Faggioli
> wrote:
> >
> > Another thing that you should check, and probably quickly mention
> > too,
> > is whether or not the callers of these functions --the ones
On Thu, 11 Feb 2016, Konrad Rzeszutek Wilk wrote:
> in the keyhandler.h file. Otherwise on ARM builds if we
> just use the keyhandler file - the compile will fail.
>
> CC: ian.campb...@citrix.com
> CC: wei.l...@citrix.com
> CC: stefano.stabell...@citrix.com
> Signed-off-by: Konrad Rzeszutek Wilk
flight 82125 xen-unstable-smoke real [real]
http://logs.test-lab.xenproject.org/osstest/logs/82125/
Failures :-/ but no regressions.
Tests which did not succeed, but are not blocking:
test-amd64-amd64-libvirt 12 migrate-support-checkfail never pass
test-armhf-armhf-xl 12
On 11.02.2016 15:13, Juergen Gross wrote:
> On 11/02/16 13:33, Daniel Kiper wrote:
>> On Thu, Feb 11, 2016 at 08:53:24AM +0100, Juergen Gross wrote:
>>> Modern pvops linux kernels support an initrd not covered by the initial
>>> mapping. This capability is flagged by an elf-note.
>>>
>>> In case
On Fri, Feb 12, 2016 at 05:00:40PM +0530, Harmandeep Kaur wrote:
> Check the return value of xc_version() and return NULL if it
> fails. libxl_get_version_info() can also return NULL now.
> Callers of the function libxl_get_version_info() are already
> prepared to deal with returning NULL on
On Thu, 11 Feb 2016, Konrad Rzeszutek Wilk wrote:
> Otherwise any code that tries to use Elf_* macros instead of
> Elf32_ or Elf_64 fails to compile.
>
> CC: ian.campb...@citrix.com
> CC: wei.l...@citrix.com
> CC: stefano.stabell...@citrix.com
> Signed-off-by: Konrad Rzeszutek Wilk
flight 81790 qemu-upstream-4.5-testing real [real]
http://logs.test-lab.xenproject.org/osstest/logs/81790/
Regressions :-(
Tests which did not succeed and are blocking,
including tests which could not be run:
test-amd64-i386-xl-qemuu-debianhvm-amd64 9 debian-hvm-install fail REGR. vs.
77858
On Feb 12, 2016 03:41, "Jan Beulich" wrote:
>
> >>> On 12.02.16 at 11:19, wrote:
> > On 02/12/2016 11:57 AM, Jan Beulich wrote:
> > On 12.02.16 at 01:22, wrote:
> >>> --- a/xen/arch/x86/vm_event.c
> >>> +++
On 12/02/16 10:57, Jan Beulich wrote:
On 12.02.16 at 11:50, wrote:
>> On 12/02/16 10:12, Jan Beulich wrote:
>> On 12.02.16 at 11:02, wrote:
On 12/02/16 10:00, Jan Beulich wrote:
On 12.02.16 at 10:51,
On Fri, 2016-02-12 at 16:34 +0530, Harmandeep Kaur wrote:
> On Fri, Feb 12, 2016 at 4:30 PM, Dario Faggioli
> wrote:
> >
> Sorry I actually meant,
> tools/libxl/xl_cmdimpl.c:vinfo = libxl_get_version_info(ctx);
>
> vinfo = libxl_get_version_info(ctx);
> if
On Thu, 11 Feb 2016, Rafael J. Wysocki wrote:
> On Thursday, February 11, 2016 04:04:14 PM Stefano Stabellini wrote:
> > On Wed, 10 Feb 2016, Rafael J. Wysocki wrote:
> > > On Tuesday, February 09, 2016 11:19:02 AM Stefano Stabellini wrote:
> > > > On Mon, 8 Feb 2016, Rafael J. Wysocki wrote:
> >
On Thu, Feb 11, 2016 at 3:18 PM, Dario Faggioli
wrote:
> Hi Harmandeep,
>
> So, I think the code in this patch is ok. Still, a few comments...
>
> On Thu, 2016-02-11 at 14:00 +0530, Harmandeep Kaur wrote:
>> Avoid handling issue of the return value of xc_version() in
From: Paul Durrant
Date: Fri, 12 Feb 2016 10:13:17 +
> This patch series adds support for frontend-configurable toeplitz hashing
> in xen-netback (on the guest receive side). This support has been testing
> against a Windows frontend and has proven to be sufficient
On Fri, Feb 12, 2016 at 4:30 PM, Dario Faggioli
wrote:
> On Fri, 2016-02-12 at 16:22 +0530, Harmandeep Kaur wrote:
>> On Thu, Feb 11, 2016 at 3:18 PM, Dario Faggioli
>> wrote:
>> >
>> > Another thing that you should check, and probably
On Fri, 12 Feb 2016, Lars Kurth wrote:
> I noticed that the qemu(1) links in
> http://xenbits.xen.org/docs/unstable/man/xl.cfg.5.html (and probably other
> docs) are broken. They go to http://man.he.net/man1/qemu, which does not
> exist any more
It doesn't look like QEMU maintains a
On Fri, 12 Feb 2016, Jan Beulich wrote:
> >>> On 12.02.16 at 12:37, wrote:
> > On Thu, 11 Feb 2016, Konrad Rzeszutek Wilk wrote:
> >> --- a/xen/include/xen/keyhandler.h
> >> +++ b/xen/include/xen/keyhandler.h
> >> @@ -19,6 +19,7 @@
> >> */
> >> typedef void
On Fri, 2016-02-12 at 12:31 +, Wei Liu wrote:
> On Fri, Feb 12, 2016 at 05:00:40PM +0530, Harmandeep Kaur wrote:
> >
> > +info->xen_version_major = r >> 16;
> > +info->xen_version_minor = r & 0xFF;
> >
> > -xc_version(ctx->xch, XENVER_extraversion, _extra);
> > +r =
>>> On 12.02.16 at 11:50, wrote:
> On 12/02/16 10:12, Jan Beulich wrote:
> On 12.02.16 at 11:02, wrote:
>>> On 12/02/16 10:00, Jan Beulich wrote:
>>> On 12.02.16 at 10:51, wrote:
> On 12/02/16 08:23,
Avoid leaking the memory mapping of the trace buffer
Coverity ID 1351228
Signed-off-by: Harmandeep Kaur
Reviewed-by: Dario Faggioli
---
v2: call to unmapping function reduced to one from two
v3: passed correct argument
> -Original Message-
> From: netdev-ow...@vger.kernel.org [mailto:netdev-
> ow...@vger.kernel.org] On Behalf Of David Miller
> Sent: 12 February 2016 10:54
> To: Paul Durrant
> Cc: net...@vger.kernel.org; xen-de...@lists.xenproject.org
> Subject: Re: [PATCH net-next v1 0/8] xen-netback:
Check the return value of xc_version() and return NULL if it
fails. libxl_get_version_info() can also return NULL now.
Callers of the function libxl_get_version_info() are already
prepared to deal with returning NULL on failure of xc_version().
Coverity ID 1351217
Signed-off-by: Harmandeep Kaur
On Fri, Feb 12, 2016 at 11:04:34AM +0200, Razvan Cojocaru wrote:
> xc_mem_access_enable_emulate() and xc_mem_access_disable_emulate()
> are currently no-ops, that is all they do is set a flag that
> nobody else checks. The user can already set the EMULATE flags in
> the vm_event response if
On 11.02.2016 13:53, Juergen Gross wrote:
> On 11/02/16 13:27, Daniel Kiper wrote:
>> On Thu, Feb 11, 2016 at 08:53:23AM +0100, Juergen Gross wrote:
>>> Do the allocation of page tables in a separate function. This will
>>> allow to do the allocation at different times of the boot preparations
>>>
>>> On 12.02.16 at 12:37, wrote:
> On Thu, 11 Feb 2016, Konrad Rzeszutek Wilk wrote:
>> --- a/xen/include/xen/keyhandler.h
>> +++ b/xen/include/xen/keyhandler.h
>> @@ -19,6 +19,7 @@
>> */
>> typedef void (keyhandler_fn_t)(unsigned char key);
>>
>> +struct
On Fri, Feb 12, 2016 at 04:38:32PM +0530, Harmandeep Kaur wrote:
> Avoid leaking the memory mapping of the trace buffer
>
> Coverity ID 1351228
>
> Signed-off-by: Harmandeep Kaur
> Reviewed-by: Dario Faggioli
Acked-by: Wei Liu
On Fri, Feb 12, 2016 at 12:50 PM, Stefano Stabellini
wrote:
> On Thu, 11 Feb 2016, Rafael J. Wysocki wrote:
>> On Thursday, February 11, 2016 04:04:14 PM Stefano Stabellini wrote:
>> > On Wed, 10 Feb 2016, Rafael J. Wysocki wrote:
>> > > On Tuesday, February 09,
On Feb 12, 2016 02:12, "Jan Beulich" wrote:
>
> >>> On 12.02.16 at 01:22, wrote:
> > Sending the dr7 register during vm_events is useful for various
applications,
> > but the current way the register value is gathered is incorrent. In this
> > patch
> >
On Fri, Feb 12, 2016 at 11:26:10AM +, Stefano Stabellini wrote:
> On Thu, 11 Feb 2016, Konrad Rzeszutek Wilk wrote:
> > Otherwise any code that tries to use Elf_* macros instead of
> > Elf32_ or Elf_64 fails to compile.
> >
> > CC: ian.campb...@citrix.com
> > CC: wei.l...@citrix.com
> > CC:
>>> On 12.02.16 at 15:06, wrote:
> On Fri, Feb 12, 2016 at 01:51:28AM -0700, Jan Beulich wrote:
>> >>> On 12.02.16 at 04:08, wrote:
>> > in the keyhandler.h file. Otherwise on ARM builds if we
>> > just use the keyhandler file - the compile will
>>> On 12.02.16 at 15:17, wrote:
> --- a/xen/include/asm-arm/config.h
> +++ b/xen/include/asm-arm/config.h
> @@ -15,8 +15,10 @@
>
> #if defined(CONFIG_ARM_64)
> # define LONG_BYTEORDER 3
> +# define ELFSIZE 64
> #else
> # define LONG_BYTEORDER 2
> +# define ELFSIZE
>>> On 12.02.16 at 15:59, wrote:
> Coverity correctly identifies that the changes in mtrr_attrib_to_str()
> introduce dead code. strings[] is a 2d array, rather than an array of
> strings, which means that strings[x] will never be a NULL pointer.
>
> Adjust the check
CC'ing other tools maintainer.
On Thu, Feb 11, 2016 at 11:37:49AM +0100, Olaf Hering wrote:
> How should libxl__initiate_device_generic_remove deal with devices which
I think you meant libxl__initiate_device_remove. There is no function
called libxl__initiate_device_generic _remove.
> have no
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
I agree to the conditions in the XenProject Coverity contribution
guidelines [0].
I'm a developer working for Citrix Systems UK, Ltd. I've been active
in the Xen community since 2008; I currently maintain Xen on ARM, Xen
support for the ARM and
>>> On 12.02.16 at 13:46, wrote:
> On Fri, 12 Feb 2016, Jan Beulich wrote:
>> and then this is the "include everything everywhere" attitude
>> which tends to needlessly slow down builds (avoiding the need
>> to include everything everywhere is actually one of the
Coverity correctly identifies that the changes in mtrr_attrib_to_str()
introduce dead code. strings[] is a 2d array, rather than an array of
strings, which means that strings[x] will never be a NULL pointer.
Adjust the check to compenstate, by looking for a NUL in strings[x][0]
instead.
On 2/11/16 9:08 PM, Konrad Rzeszutek Wilk wrote:
> c/s 361b4f9f0f0d4adc19df428e224a7b8fa62cd392
> "build: save generated xen .config" forgot to remove
> the config file when uninstalling.
>
> CC: ian.campb...@citrix.com
> CC: jbeul...@suse.com
> CC: car...@cardoe.com
> CC:
[Yes, replying to myself]
On Fri, 2016-02-12 at 11:50 +0100, Dario Faggioli wrote:
> On Fri, 2016-02-12 at 02:50 -0700, Jan Beulich wrote:
> > > > > On 12.02.16 at 10:37, wrote:
> > > @@ -787,6 +788,16 @@ _csched_cpu_pick(const struct scheduler
> > > *ops,
> > > struct
On 12/02/16 14:53, Stefano Stabellini wrote:
> I agree to the conditions in the XenProject Coverity contribution
> guidelines [0].
>
> I'm a developer working for Citrix Systems UK, Ltd. I've been active
> in the Xen community since 2008; I currently maintain Xen on ARM, Xen
> support for the ARM
On Fri, Feb 12, 2016 at 01:51:28AM -0700, Jan Beulich wrote:
> >>> On 12.02.16 at 04:08, wrote:
> > in the keyhandler.h file. Otherwise on ARM builds if we
> > just use the keyhandler file - the compile will fail.
> >
> > CC: ian.campb...@citrix.com
> > CC:
On 12/02/16 13:24, Vladimir 'φ-coder/phcoder' Serbinenko wrote:
> On 11.02.2016 15:13, Juergen Gross wrote:
>> On 11/02/16 13:33, Daniel Kiper wrote:
>>> On Thu, Feb 11, 2016 at 08:53:24AM +0100, Juergen Gross wrote:
Modern pvops linux kernels support an initrd not covered by
the initial
flight 38740 distros-debian-jessie real [real]
http://osstest.xs.citrite.net/~osstest/testlogs/logs/38740/
Failures :-/ but no regressions.
Tests which did not succeed, but are not blocking:
test-armhf-armhf-armhf-jessie-netboot-pygrub 12 saverestore-support-check fail
never pass
On Fri, Feb 12, Wei Liu wrote:
> CC'ing other tools maintainer.
>
> On Thu, Feb 11, 2016 at 11:37:49AM +0100, Olaf Hering wrote:
> > How should libxl__initiate_device_generic_remove deal with devices which
>
> I think you meant libxl__initiate_device_remove. There is no function
> called
>>> On 12.02.16 at 13:50, wrote:
> On Feb 12, 2016 03:41, "Jan Beulich" wrote:
>> In which case ASSERT(is_hvm_vcpu(curr)) would be the common
>> way to document this (at once avoiding the open coding of
>> is_hvm_vcpu()).
>>
>
> I don't think we need an
flight 81798 qemu-upstream-4.6-testing real [real]
http://logs.test-lab.xenproject.org/osstest/logs/81798/
Regressions :-(
Tests which did not succeed and are blocking,
including tests which could not be run:
test-amd64-amd64-xl-qemuu-debianhvm-amd64 9 debian-hvm-install fail REGR. vs.
77722
Applied, thanks
On 20.07.2015 16:35, Daniel Kiper wrote:
> malloc_in_range() should not use memory region if its starta is smaller
> than size. Otherwise target wraps around and points to region which is
> usually not a RAM, e.g.:
>
> loader/multiboot.c:93: segment 0: paddr=0x80,
>>> On 12.02.16 at 01:22, wrote:
> --- a/xen/arch/x86/hvm/event.c
> +++ b/xen/arch/x86/hvm/event.c
> @@ -23,40 +23,9 @@
> #include
> #include
> #include
> +#include
> #include
>
> -static void hvm_event_fill_regs(vm_event_request_t *req)
> -{
> -const struct
>>> On 12.02.16 at 01:22, wrote:
> --- a/xen/arch/x86/vm_event.c
> +++ b/xen/arch/x86/vm_event.c
> @@ -122,6 +122,65 @@ void vm_event_set_registers(struct vcpu *v,
> vm_event_response_t *rsp)
> v->arch.user_regs.eip = rsp->data.regs.x86.rip;
> }
>
> +void
On 12/02/16 09:16, Jan Beulich wrote:
On 11.02.16 at 20:51, wrote:
>> Otherwise, debug code such as "void __attribute__((noreturn)) foobar()" fails
>> to compile when the noreturn itself gets expanded, resulting in
>> __attribute__((__attribute__((noreturn.
>
The canonical netif header (in the Xen source repo) and the Linux variant
have diverged significantly. Recently much documentation has been added to
the canonical header and new definitions and types to support packet hash
configuration. Subsequent patches in this series add support for packet
This patch series adds support for frontend-configurable toeplitz hashing
in xen-netback (on the guest receive side). This support has been testing
against a Windows frontend and has proven to be sufficient to pass the
Microsoft HCK NDIS RSS tests.
For convenience my development branch is
This patch adds code to xen-netback to use the value in a toeplitz hash
extra info fragment passed from the VM frontend in a transmit-side packet
to set the skb hash accordingly.
Signed-off-by: Paul Durrant
Cc: Ian Campbell
Cc: Wei Liu
My recent patch to include/xen/interface/io/netif.h defines a new extra
info type that can be used to pass toeplitz hash values between backend
and VM frontend.
This patch adds code to xen-netback to pass hash values calculated for
receive-side packets to the VM frontend.
Signed-off-by: Paul
My recent patch to include/xen/interface/io/netif.h defines a new shared
ring (in addition to the rx and tx rings) for passing control messages
from a VM frontend driver to a backend driver.
This patch adds the necessary code to xen-netback to map this new shared
ring, should it be created by a
...for receive-side packets.
My recent patch to include/xen/interface/io/netif.h defines a set of
control messages that can be used by a VM frontend driver to configure
toeplitz hashing of receive-side packets and consequent steering of those
packets to particular queues.
This patch introduces
...rather than a boolean merely indicating a canonical L4 hash.
skb_set_hash() takes a hash type (from enum pkt_hash_types) as an
argument but information is lost since only a single bit in the skb
stores whether that hash type is PKT_HASH_TYPE_L4 or not.
By using two bits it's possible to store
>>> On 05.02.16 at 14:42, wrote:
> --- /dev/null
> +++ b/xen/arch/x86/cpuid.c
> @@ -0,0 +1,19 @@
> +#include
> +#include
> +
> +const uint32_t known_features[] = INIT_KNOWN_FEATURES;
> +
> +static void __maybe_unused build_assertions(void)
> +{
> +
On 12/02/16 16:36, Jan Beulich wrote:
On 05.02.16 at 14:41, wrote:
>> This script consumes include/public/arch-x86/cpufeatureset.h and generates a
>> single include/asm-x86/cpuid-autogen.h containing all the processed
>> information.
>>
>> Signed-off-by: Andrew
>>> On 05.02.16 at 14:42, wrote:
> Awkwardly, some new feature bits mean "Feature $X no longer works".
> Store these inverted in a featureset.
>
> This permits safe zero-extending of a smaller featureset as part of a
> comparison, and safe reasoning (subset?,
On 12/02/16 16:47, Jan Beulich wrote:
On 05.02.16 at 14:42, wrote:
>> Awkwardly, some new feature bits mean "Feature $X no longer works".
>> Store these inverted in a featureset.
>>
>> This permits safe zero-extending of a smaller featureset as part of a
>>
>>> On 05.02.16 at 14:42, wrote:
> Use attributes to specify whether a feature is applicable to be exposed to:
> 1) All guests
> 2) HVM guests
> 3) HVM HAP guests
No provisions for PV-only or shadow-only features?
> +#define X86_FEATURE_MTRR ( 0*32+12) /*S
>>> On 12.02.16 at 17:48, wrote:
> On 12/02/16 16:43, Jan Beulich wrote:
> On 05.02.16 at 14:42, wrote:
>>> --- /dev/null
>>> +++ b/xen/arch/x86/cpuid.c
>>> @@ -0,0 +1,19 @@
>>> +#include
>>> +#include
>>> +
>>> +const uint32_t
Hey,
Fedora for the longest time seems to have two linkers - one normal for GNU
applications and then another - mingw64 - for building EFI applications.
Which means that to compile ELF binaries on Fedora requires this patch
(taken from Fedora build):
>From
The following changes since commit f075c89f0a9cb31daf38892371d2822177505706:
Merge remote-tracking branch 'remotes/stefanha/tags/block-pull-request' into
staging (2016-02-09 17:56:46 +)
are available in the git repository at:
git://xenbits.xen.org/people/sstabellini/qemu-dm.git
On Fri, Feb 12, 2016 at 03:53:06PM +0100, Olaf Hering wrote:
> On Fri, Feb 12, Wei Liu wrote:
>
> > CC'ing other tools maintainer.
> >
> > On Thu, Feb 11, 2016 at 11:37:49AM +0100, Olaf Hering wrote:
> > > How should libxl__initiate_device_generic_remove deal with devices which
> >
> > I think
> -Original Message-
> From: David Miller [mailto:da...@davemloft.net]
> Sent: 12 February 2016 16:42
> To: Paul Durrant
> Cc: net...@vger.kernel.org; xen-de...@lists.xenproject.org
> Subject: Re: [PATCH net-next v1 0/8] xen-netback: support toeplitz hashing
>
> From: Paul Durrant
Doug Goldstein writes ("[PATCH v3] travis: add initial Travis CI script to do
builds"):
> This is just suppose to do a simple compile test on Travis CI. Currently
> due to linux86 (bcc/bin86/dev86) not being whitelisted the tools cannot
> be built.
>
> Signed-off-by: Doug Goldstein
On Thu, Feb 11, 2016 at 03:43:30PM +, Olaf Hering wrote:
> Signed-off-by: Olaf Hering
> Acked-by: Ian Campbell
> Cc: Ian Campbell
> Cc: Ian Jackson
> Cc: Jan Beulich
> Cc: Keir
On Thu, Feb 11, 2016 at 03:43:27PM +, Olaf Hering wrote:
> The pvops kernel expects either "naa.WWN:LUN" or "h:c:t:l" in the p-dev
> property. Add the missing :LUN part to the comment.
>
> Signed-off-by: Olaf Hering
> Acked-by: Ian Campbell
> Cc: Ian
On Thu, Feb 11, 2016 at 03:43:31PM +, Olaf Hering wrote:
> Just to make them public, not meant for merging:
I might be mistaken, but if you don't provide a hotplug script or some
sort for Xen how do you expect user to make use vscsi?
Wei.
___
On 12/02/16 16:43, Jan Beulich wrote:
On 05.02.16 at 14:42, wrote:
>> --- /dev/null
>> +++ b/xen/arch/x86/cpuid.c
>> @@ -0,0 +1,19 @@
>> +#include
>> +#include
>> +
>> +const uint32_t known_features[] = INIT_KNOWN_FEATURES;
>> +
>> +static void __maybe_unused
>>> On 12.02.16 at 17:50, wrote:
> On 12/02/16 16:47, Jan Beulich wrote:
> On 05.02.16 at 14:42, wrote:
>>> Awkwardly, some new feature bits mean "Feature $X no longer works".
>>> Store these inverted in a featureset.
>>>
>>> This permits
I notice you posted several enquiries in your previous patch, but I'm
not sure if any of them still requires answers because you seemed to
have posted answers to your own enquiries. I skip all of them and start
from this version.
There seems to be a lot of places in this patch where the lines are
1 - 100 of 184 matches
Mail list logo