flight 100590 ovmf real [real]
http://logs.test-lab.xenproject.org/osstest/logs/100590/
Perfect :-)
All tests in this flight passed as required
version targeted for testing:
ovmf 90fadfc00ad773441aeb97718cd9c2e3dc7d2f71
baseline version:
ovmf 72388f9c10126a718a7ac
>>> On 22.08.16 at 19:24, wrote:
> On 04/08/16 17:02, Jan Beulich wrote:
> On 04.08.16 at 17:16, wrote:
>>> Using mlock() for hypercall buffers is not sufficient since mlocked
>>> pages are still subject to compaction and page migration. Page
>>> migration can be prevented by taking addition
On 23/08/16 09:46, Jan Beulich wrote:
On 22.08.16 at 19:24, wrote:
>> On 04/08/16 17:02, Jan Beulich wrote:
>> On 04.08.16 at 17:16, wrote:
Using mlock() for hypercall buffers is not sufficient since mlocked
pages are still subject to compaction and page migration. Page
m
>>> On 22.08.16 at 19:30, wrote:
> On 19/08/16 19:07, Andrew Cooper wrote:
>> On 19/08/16 18:09, Andrew Cooper wrote:
>>> On 19/08/16 13:53, Jan Beulich wrote:
User mode code generally cannot be expected to invoke the PV-enabled
CPUID Xen supports, and prior to the CPUID levelling change
On 23/08/16 10:00, Jan Beulich wrote:
On 22.08.16 at 19:30, wrote:
>> On 19/08/16 19:07, Andrew Cooper wrote:
>>> On 19/08/16 18:09, Andrew Cooper wrote:
On 19/08/16 13:53, Jan Beulich wrote:
> User mode code generally cannot be expected to invoke the PV-enabled
> CPUID Xen suppo
On 22/08/16 16:05, Wei Liu wrote:
> They shouldn't be available when the respective BIOS is disabled at build
> time.
>
> This should fix the issus that causes xtf fail to launch hvm guests.
>
> Wei Liu (2):
> tools: only define {OVMF,SEABIOS}_PATH when they are enabled
> libxl: only return {O
>>> On 23.08.16 at 11:24, wrote:
> On 23/08/16 10:00, Jan Beulich wrote:
> On 22.08.16 at 19:30, wrote:
>>> On 19/08/16 19:07, Andrew Cooper wrote:
On 19/08/16 18:09, Andrew Cooper wrote:
> On 19/08/16 13:53, Jan Beulich wrote:
>> User mode code generally cannot be expected to in
On 23/08/16 10:41, Jan Beulich wrote:
On 23.08.16 at 11:24, wrote:
>> On 23/08/16 10:00, Jan Beulich wrote:
>> On 22.08.16 at 19:30, wrote:
On 19/08/16 19:07, Andrew Cooper wrote:
> On 19/08/16 18:09, Andrew Cooper wrote:
>> On 19/08/16 13:53, Jan Beulich wrote:
>>> User
>>> On 23.08.16 at 11:48, wrote:
> On 23/08/16 10:41, Jan Beulich wrote:
> On 23.08.16 at 11:24, wrote:
>>> On 23/08/16 10:00, Jan Beulich wrote:
>>> On 22.08.16 at 19:30, wrote:
> On 19/08/16 19:07, Andrew Cooper wrote:
>> On 19/08/16 18:09, Andrew Cooper wrote:
>>> On 19/08
On Mon, Aug 22, 2016 at 08:33:11PM +0200, Mikko Rapeli wrote:
> xen/interface/xen.h is not exported from kernel headers so remove the
> dependency and provide needed defines for domid_t and xen_pfn_t if they
> are not already defined by some other e.g. Xen specific headers.
I'm confused. How did
This run is configured for baseline tests only.
flight 67582 ovmf real [real]
http://osstest.xs.citrite.net/~osstest/testlogs/logs/67582/
Perfect :-)
All tests in this flight passed as required
version targeted for testing:
ovmf 90fadfc00ad773441aeb97718cd9c2e3dc7d2f71
baseline v
flight 100589 qemu-mainline real [real]
http://logs.test-lab.xenproject.org/osstest/logs/100589/
Failures :-/ but no regressions.
Regressions which are regarded as allowable (not blocking):
test-amd64-i386-xl-qemuu-win7-amd64 16 guest-stop fail like 100581
test-amd64-amd64-xl-rtds
This run is configured for baseline tests only.
flight 67580 xen-unstable real [real]
http://osstest.xs.citrite.net/~osstest/testlogs/logs/67580/
Failures :-/ but no regressions.
Regressions which are regarded as allowable (not blocking):
build-amd64-rumpuserxen 6 xen-build
flight 100593 libvirt real [real]
http://logs.test-lab.xenproject.org/osstest/logs/100593/
Failures :-/ but no regressions.
Tests which did not succeed, but are not blocking:
test-amd64-amd64-libvirt-xsm 12 migrate-support-checkfail never pass
test-amd64-i386-libvirt-xsm 12 migrate-s
flight 67581 distros-debian-snapshot real [real]
http://osstest.xs.citrite.net/~osstest/testlogs/logs/67581/
Failures :-/ but no regressions.
Regressions which are regarded as allowable (not blocking):
test-armhf-armhf-armhf-daily-netboot-pygrub 9 debian-di-install fail like 67540
test-amd64-i3
flight 100592 xen-unstable real [real]
http://logs.test-lab.xenproject.org/osstest/logs/100592/
Failures :-/ but no regressions.
Tests which are failing intermittently (not blocking):
test-amd64-amd64-xl-rtds 6 xen-boot fail in 100585 pass in 100592
test-armhf-armhf-xl-arndale 6
Trap handling in HVMlite domain is different from pv one.
Signed-off-by: Juergen Gross
---
arch/x86/traps.c | 2 --
arch/x86/x86_32.S | 19 ++--
arch/x86/x86_64.S | 22 +-
include/x86/os.h | 67 +++
4 files change
Select what to print in arch_init() depending on the domain type.
Signed-off-by: Juergen Gross
---
arch/x86/setup.c | 47 ---
1 file changed, 32 insertions(+), 15 deletions(-)
diff --git a/arch/x86/setup.c b/arch/x86/setup.c
index 50aa504..948b08a 100
A HVMlite domain is always starting in 32 bit mode. Add the appropriate
boot code to arch/x86 for the non-paravirtualized case.
For this boot code to become active we need to suppress the pv related
elfnotes and add an appropriate elfnote for HVMlite.
As the HVMlite boot code is more or less the
The command line parameters are passed in start_info structure for pv,
but via hvm_start_info for HVMlite. Add a generic function to fill a
local cmdline variable with the command line string.
Signed-off-by: Juergen Gross
---
arch/x86/setup.c | 20 ++--
include/kernel.h | 3 +++
When running in HVMlite mode we need to setup the hypercall page by
ourself.
Signed-off-by: Juergen Gross
---
arch/x86/events.c | 4 ++--
arch/x86/setup.c | 26 ++
include/x86/os.h | 22 +-
3 files changed, 45 insertions(+), 7 deletions(-)
diff --g
This patch series adds HVMlite support for Mini-OS. Setting
CONFIG_PARAVIRT to "n" (default is "y") will build mini-os as a
HVMlite domain on x86. Both 32- and 64-bit builds are supported.
Tested with plain Mini-OS configuration to work in 32- and 64-bit
mode either paravirtualized or as HVM domai
The start_info structure should be used only in case of CONFIG_PARAVIRT
defined. Remove it from being used in other places. Especially the
usage as parameter for applications linked to the kernel is no good
idea.
Signed-off-by: Juergen Gross
---
arch/arm/setup.c | 8
arch/x86/setup
Add a service function to map the shared info page on a
non-paravirtualized system. The code is already existing on ARM side,
just move it to hypervisor.c.
Signed-off-by: Juergen Gross
---
arch/arm/arm32.S | 4 ++--
arch/arm/setup.c | 11 +--
arch/x86/setup.c | 38 ++
For support of HVMlite don't use mmu_update hypercalls, but write the
page table entries directly.
Signed-off-by: Juergen Gross
---
arch/x86/mm.c | 147 +-
arch/x86/traps.c | 10
include/x86/arch_mm.h | 4 ++
include/x86/os.h
Trap handling for HVMlite domains requires an initialized IDT.
Signed-off-by: Juergen Gross
---
arch/x86/setup.c | 15 --
arch/x86/traps.c | 88 +++
arch/x86/x86_64.S | 4 +++
include/x86/os.h | 1 +
4 files changed, 93 insertions(
In order to support HVMlite we need to get the ring page and event
channel from the hypervisor via hypercalls. Move the already existing
get_xenbus() function from arm specific coding to xenbus/xenbus.c and
provide a similar paravirtualized function.
Signed-off-by: Juergen Gross
---
arch/arm/set
Especially page table entry definitions are currently not usable from
assembler sources on x86 as the constants are defined with ULL suffix.
Change this by adding the suffix only when the header is included from
a C source.
Hide some C prototypes when in assembler environment.
Signed-off-by: Juer
dump_regs() will result in page fault in early boot as there is no
current thread pointer. Handle this case.
Signed-off-by: Juergen Gross
---
arch/x86/traps.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/arch/x86/traps.c b/arch/x86/traps.c
index 6353718..307a14c 100644
---
Use the latest Xen headers.
Signed-off-by: Juergen Gross
---
include/xen/arch-x86/hvm/start_info.h | 98 +++
include/xen/elfnote.h | 12 -
2 files changed, 109 insertions(+), 1 deletion(-)
create mode 100644 include/xen/arch-x86/hvm/start_info
Instead of calling HYPERVISOR_sched_op(SCHEDOP_block, 0) we need to use
the "hlt" instruction with interrupts enabled to switch to idle when
running as HVMlite domain.
This requires to setup a new timer in the timer handler as there is no
guarantee the original timer event we are waiting for won't
In order to support HVMlite we need to get the ring page and event
channel from the hypervisor via hypercalls. Move the already existing
get_console() function from arm specific coding to
console/xencons_ring.c and provide a similar paravirtualized function.
Signed-off-by: Juergen Gross
---
arch
Add a new config macro CONFIG_PARAVIRT which defaults to be defined on
x86. This is the first step for supporting a HVMlite Mini-OS.
Doing this via CONFIG_PARAVIRT instead of something like CONFIG_HVMLITE
was chosen as the arm port can then drop some dummy routines needed for
para-virtualization o
On HVMlite we have to look at the physical memory map to know which
memory frames are usable.
In order to make life easier we define a dummy memory map for other
domain types (pv and arm) which has just one entry with a maximum
memory size.
Signed-off-by: Juergen Gross
---
arch/arm/mm.c
Add a function map_frame_virt() to map a given frame and return its
virtual address.
On arm we just use the frame physical address, while on x86 we take a
page from the virtual kernel area. For this purpose make this area
available even in case of undefined CONFIG_BALLOON.
Signed-off-by: Juergen
Support the HYPERVISOR_hvm_op() hypercall which is needed for HVMlite.
Add convenience functions hvm_get_parameter() and hvm_set_parameter().
Signed-off-by: Juergen Gross
---
arch/arm/setup.c | 15 ---
hypervisor.c | 27 ++
All handling related to p2m is needed for paravirtualized environment
only. So put all functions operating on p2m list in #ifdef
CONFIG_PARAVIRT sections. Add a paravirt.h header defining dummy
functions for non-paravirtualized environments. Instead of using
references to start_info use dedicated v
In HVMlite mode we need the virtual kernel area for mapping of the
console and xenbus ring pages as especially on 32 bit architecture
their pfns might be above the supported maximum memory size.
Add the page tables needed for doing the mapping.
Signed-off-by: Juergen Gross
---
arch/x86/mm.c
When remapping the page allocator's bitmap for the ballooning support
the calculation of the needed size is wrong. This doesn't really
matter today as nothing is allocated after that bitmap, but it should
be corrected nevertheless.
Signed-off-by: Juergen Gross
---
balloon.c | 4 ++--
1 file chan
On Tue, Aug 23, 2016 at 12:11:40AM +0900, Masami Hiramatsu wrote:
> On Fri, 19 Aug 2016 14:34:12 -0700
> mcg...@kernel.org wrote:
>
> > From: "Luis R. Rodriguez"
> >
> > Often all is needed is these small helpers, instead of compiler.h
> > or a full kprobes.h. This is important for asm helpers,
flight 100594 linux-4.1 real [real]
http://logs.test-lab.xenproject.org/osstest/logs/100594/
Failures :-/ but no regressions.
Tests which are failing intermittently (not blocking):
test-amd64-i386-libvirt-qemuu-debianhvm-amd64-xsm 9 debian-hvm-install fail in
100587 pass in 100594
test-armhf-a
A subsequent change needs to special-case OSXSAVE handling, which is per-vcpu
rather than per-domain.
No functional change.
Signed-off-by: Andrew Cooper
---
CC: Jan Beulich
---
xen/arch/x86/cpu/amd.c | 3 ++-
xen/arch/x86/cpu/common.c | 4 ++--
xen/arch/x86/cpu/intel.c|
Contrary to c/s b2507fe7 "x86/domctl: Update PV domain cpumasks when setting
cpuid policy", Intel CPUID masks are applied after fast forwarding hardware
state, rather than before. (All behaviour in this regard appears completely
undocumented by both Intel and AMD).
Therefore, a set bit in the MSR
There is no need to extend the workaround to the faulted CPUID view, as
Linux's dependence on the workaround is stricly via the emulated view.
This causes a guest kernel faulted CPUID to observe architectural behaviour
with respect to its CR4.OSXSAVE setting.
Signed-off-by: Andrew Cooper
---
CC:
This run is configured for baseline tests only.
flight 67583 qemu-mainline real [real]
http://osstest.xs.citrite.net/~osstest/testlogs/logs/67583/
Failures :-/ but no regressions.
Regressions which are regarded as allowable (not blocking):
test-amd64-amd64-qemuu-nested-intel 16 debian-hvm-insta
On Tue, Aug 23, 2016 at 11:26:33AM +1000, Nicholas Piggin wrote:
> On Fri, 19 Aug 2016 14:34:02 -0700
> mcg...@kernel.org wrote:
> > +/**
> > + * DOC: Standard ELF section use in Linux
> > + *
> > + * Linux makes use of the standard ELF sections, this sections documents
> > + * these.
> > + */
> >
flight 100598 ovmf real [real]
http://logs.test-lab.xenproject.org/osstest/logs/100598/
Perfect :-)
All tests in this flight passed as required
version targeted for testing:
ovmf 93e59f76fe502100a66e971acfbc1c58c63f564d
baseline version:
ovmf 90fadfc00ad773441aeb9
On Tue, 23 Aug 2016, Russell King - ARM Linux wrote:
> On Mon, Aug 22, 2016 at 08:33:11PM +0200, Mikko Rapeli wrote:
> > xen/interface/xen.h is not exported from kernel headers so remove the
> > dependency and provide needed defines for domid_t and xen_pfn_t if they
> > are not already defined by s
flight 100597 linux-3.18 real [real]
http://logs.test-lab.xenproject.org/osstest/logs/100597/
Failures :-/ but no regressions.
Tests which are failing intermittently (not blocking):
test-armhf-armhf-xl-xsm 15 guest-start/debian.repeat fail in 100588 pass in
100597
test-armhf-armhf-xl-multivcpu
David Vrabel writes:
> On 22/08/16 16:42, Vitaly Kuznetsov wrote:
>>
>> I see two ways to fix the issue:
>> - Change the 'wire' protocol between netfront and netback to start keeping
>> the original SKB structure. We'll have to add a flag indicating the fact
>> that the particular request is
Juergen Gross, on Tue 23 Aug 2016 17:15:48 +0200, wrote:
> dump_regs() will result in page fault in early boot as there is no
> current thread pointer. Handle this case.
>
> Signed-off-by: Juergen Gross
Reviewed-by: Samuel Thibault
> ---
> arch/x86/traps.c | 2 +-
> 1 file changed, 1 insertio
Juergen Gross, on Tue 23 Aug 2016 17:15:47 +0200, wrote:
> Use the latest Xen headers.
>
> Signed-off-by: Juergen Gross
Reviewed-by: Samuel Thibault
> ---
> include/xen/arch-x86/hvm/start_info.h | 98
> +++
> include/xen/elfnote.h | 12 -
>
Juergen Gross, on Tue 23 Aug 2016 17:15:50 +0200, wrote:
> Especially page table entry definitions are currently not usable from
> assembler sources on x86 as the constants are defined with ULL suffix.
> Change this by adding the suffix only when the header is included from
> a C source.
>
> Hide
Juergen Gross, on Tue 23 Aug 2016 17:15:59 +0200, wrote:
> When remapping the page allocator's bitmap for the ballooning support
> the calculation of the needed size is wrong. This doesn't really
> matter today as nothing is allocated after that bitmap, but it should
> be corrected nevertheless.
O
Juergen Gross, on Tue 23 Aug 2016 17:15:49 +0200, wrote:
> Add a new config macro CONFIG_PARAVIRT which defaults to be defined on
> x86. This is the first step for supporting a HVMlite Mini-OS.
>
> Doing this via CONFIG_PARAVIRT instead of something like CONFIG_HVMLITE
> was chosen as the arm port
On 8/18/16 10:13 AM, Wei Liu wrote:
>
> +if (info->device_model_version == LIBXL_DEVICE_MODEL_VERSION_QEMU_XEN) {
> +if (info->u.hvm.system_firmware) {
> +bios_filename = info->u.hvm.system_firmware;
> +} else {
> +switch (info->u.hvm.bios) {
> +
On Tue, Aug 23, 2016 at 11:13:52AM -0700, Stefano Stabellini wrote:
> On Tue, 23 Aug 2016, Russell King - ARM Linux wrote:
> > On Mon, Aug 22, 2016 at 08:33:11PM +0200, Mikko Rapeli wrote:
> > > xen/interface/xen.h is not exported from kernel headers so remove the
> > > dependency and provide neede
Hello,
It looks good, there is just:
Juergen Gross, on Tue 23 Aug 2016 17:15:51 +0200, wrote:
> +.macro PTES prot, addr, incr, idx, end
Please document what prot, addr, incr, idx, end represent precisely.
> +.ifgt \end-\idx-32
> +PTES \prot, \addr, \incr, \idx, "(\idx+32
hello,
Juergen Gross, on Tue 23 Aug 2016 17:15:52 +0200, wrote:
> When running in HVMlite mode we need to setup the hypercall page by
> ourself.
>
> Signed-off-by: Juergen Gross
> ---
>
> +#else /* ifdef __x86_64__ */
> +#error "Unsupported architecture"
> +#endif
> +
> #define wrmsr(msr,val1
flight 100601 ovmf real [real]
http://logs.test-lab.xenproject.org/osstest/logs/100601/
Perfect :-)
All tests in this flight passed as required
version targeted for testing:
ovmf 4962fcfa7d265824f01f74d782d5ed841ec8a72f
baseline version:
ovmf 93e59f76fe502100a66e9
On Tue, 23 Aug 2016, Mikko Rapeli wrote:
> On Tue, Aug 23, 2016 at 11:13:52AM -0700, Stefano Stabellini wrote:
> > On Tue, 23 Aug 2016, Russell King - ARM Linux wrote:
> > > On Mon, Aug 22, 2016 at 08:33:11PM +0200, Mikko Rapeli wrote:
> > > > xen/interface/xen.h is not exported from kernel headers
Juergen Gross, on Tue 23 Aug 2016 17:15:53 +0200, wrote:
> Support the HYPERVISOR_hvm_op() hypercall which is needed for HVMlite.
> Add convenience functions hvm_get_parameter() and hvm_set_parameter().
>
> Signed-off-by: Juergen Gross
Reviewed-by: Samuel Thibault
> ---
> arch/arm/setup.c
Juergen Gross, on Tue 23 Aug 2016 17:15:54 +0200, wrote:
> Trap handling for HVMlite domains requires an initialized IDT.
>
> Signed-off-by: Juergen Gross
Reviewed-by: Samuel Thibault
> ---
> arch/x86/setup.c | 15 --
> arch/x86/traps.c | 88
> ++
Juergen Gross, on Tue 23 Aug 2016 17:15:55 +0200, wrote:
> Trap handling in HVMlite domain is different from pv one.
>
> Signed-off-by: Juergen Gross
Reviewed-by: Samuel Thibault
> ---
> arch/x86/traps.c | 2 --
> arch/x86/x86_32.S | 19 ++--
> arch/x86/x86_64.S | 22 +++
Juergen Gross, on Tue 23 Aug 2016 17:15:56 +0200, wrote:
> All handling related to p2m is needed for paravirtualized environment
> only. So put all functions operating on p2m list in #ifdef
> CONFIG_PARAVIRT sections. Add a paravirt.h header defining dummy
> functions for non-paravirtualized enviro
Juergen Gross, on Tue 23 Aug 2016 17:15:57 +0200, wrote:
> In HVMlite mode we need the virtual kernel area for mapping of the
> console and xenbus ring pages as especially on 32 bit architecture
> their pfns might be above the supported maximum memory size.
>
> Add the page tables needed for doing
Juergen Gross, on Tue 23 Aug 2016 17:15:58 +0200, wrote:
> For support of HVMlite don't use mmu_update hypercalls, but write the
> page table entries directly.
>
> Signed-off-by: Juergen Gross
Reviewed-by: Samuel Thibault
> ---
> arch/x86/mm.c | 147
>
Juergen Gross, on Tue 23 Aug 2016 17:16:00 +0200, wrote:
> Add a function map_frame_virt() to map a given frame and return its
> virtual address.
>
> On arm we just use the frame physical address, while on x86 we take a
> page from the virtual kernel area. For this purpose make this area
> availab
Juergen Gross, on Tue 23 Aug 2016 17:16:01 +0200, wrote:
> In order to support HVMlite we need to get the ring page and event
> channel from the hypervisor via hypercalls. Move the already existing
> get_console() function from arm specific coding to
> console/xencons_ring.c and provide a similar p
Juergen Gross, on Tue 23 Aug 2016 17:16:02 +0200, wrote:
> In order to support HVMlite we need to get the ring page and event
> channel from the hypervisor via hypercalls. Move the already existing
> get_xenbus() function from arm specific coding to xenbus/xenbus.c and
> provide a similar paravirtu
Juergen Gross, on Tue 23 Aug 2016 17:16:04 +0200, wrote:
> Add a service function to map the shared info page on a
> non-paravirtualized system. The code is already existing on ARM side,
> just move it to hypervisor.c.
>
> Signed-off-by: Juergen Gross
Reviewed-by: Samuel Thibault
> ---
> arch
Juergen Gross, on Tue 23 Aug 2016 17:16:05 +0200, wrote:
> The start_info structure should be used only in case of CONFIG_PARAVIRT
> defined. Remove it from being used in other places. Especially the
> usage as parameter for applications linked to the kernel is no good
> idea.
>
> Signed-off-by: J
Juergen Gross, on Tue 23 Aug 2016 17:16:06 +0200, wrote:
> +printk("Xen Minimal OS (hvm)!\n");
> +printk(" start_info: %p(VA)\n", si);
> +printk(" shared_inf: %p(VA)\n", HYPERVISOR_shared_info);
> +printk(" modlist: 0x%lx(PA)\n", (unsigned long)si->modlist_paddr);
It would be
Juergen Gross, on Tue 23 Aug 2016 17:16:07 +0200, wrote:
> On HVMlite we have to look at the physical memory map to know which
> memory frames are usable.
>
> In order to make life easier we define a dummy memory map for other
> domain types (pv and arm) which has just one entry with a maximum
> m
Juergen Gross, on Tue 23 Aug 2016 17:16:08 +0200, wrote:
> Instead of calling HYPERVISOR_sched_op(SCHEDOP_block, 0) we need to use
> the "hlt" instruction with interrupts enabled to switch to idle when
> running as HVMlite domain.
>
> This requires to setup a new timer in the timer handler as ther
Juergen Gross, on Tue 23 Aug 2016 17:16:03 +0200, wrote:
> The command line parameters are passed in start_info structure for pv,
> but via hvm_start_info for HVMlite. Add a generic function to fill a
> local cmdline variable with the command line string.
>
> Signed-off-by: Juergen Gross
Reviewe
From: Wei Yongjun
Date: Mon, 22 Aug 2016 23:01:29 +
> From: Wei Yongjun
>
> The callback function of call_rcu() just calls a kfree(), so we
> can use kfree_rcu() instead of call_rcu() + callback function.
>
> Signed-off-by: Wei Yongjun
Applied.
__
This run is configured for baseline tests only.
flight 67584 linux-4.1 real [real]
http://osstest.xs.citrite.net/~osstest/testlogs/logs/67584/
Regressions :-(
Tests which did not succeed and are blocking,
including tests which could not be run:
test-amd64-i386-qemut-rhel6hvm-intel 9 redhat-ins
This run is configured for baseline tests only.
flight 67585 ovmf real [real]
http://osstest.xs.citrite.net/~osstest/testlogs/logs/67585/
Perfect :-)
All tests in this flight passed as required
version targeted for testing:
ovmf 93e59f76fe502100a66e971acfbc1c58c63f564d
baseline v
It is possible. Especially if the only thing they do is
NOP functions - in which case there is only .livepatch.funcs
sections.
Reviewed-by: Jan Beulich
Reviewed-by: Ross Lagerwall
Signed-off-by: Konrad Rzeszutek Wilk
---
Cc: Konrad Rzeszutek Wilk
Cc: Ross Lagerwall
Cc: Jan Beulich
Cc: Andre
Specifically the code that is looking up f->old_addr - which
can be in its own routine instead of having it part of prepare_payload.
No functional change.
Signed-off-by: Konrad Rzeszutek Wilk
---
Cc: Ross Lagerwall
v3: First submission.
v4: Added const to last parameter of livepatch_symbols_
From: Ross Lagerwall
Add hook functions which run during patch apply and patch revert.
Hook functions are used by livepatch payloads to manipulate data
structures during patching, etc.
One use case is the XSA91. As Martin mentions it:
"If we have shadow variables, we also need an unload hook to
Livepatch expected at some point to be able to print the
build-id during bootup, which it did not. The reason is
that xen_build_init and livepatch_init are both __initcall
type routines. This meant that when livepatch_init called
xen_build_id, it would return -ENODATA as build_id_len was
not setup
Hey!
Since v3: [https://lists.xen.org/archives/html/xen-devel/2016-08/msg01825.html]
- Acked on reviews
v2, v1:
- Left over fixes and features that didn't get quite done in 4.7
Included are:
- Bug-fixes
- Parsing of symbol names encoded as: symbol+0x
- NOP patching
- Generating an symbol ma
The NOP functionality will NOP any of the code at
the 'old_addr' or at 'name' if the 'new_addr' is zero.
The purpose of this is to NOP out calls, such as:
e8 <4-bytes-offset>
(5 byte insn), or on ARM a 4 byte insn for branching.
But on x86 we could NOP instructions that are much
shorter or longe
You could construct _most_ of the names of the functions
by doing 'nm --defined' but unfortunatly you do not get the
prefix that is added on in Xen . For example:
$ cat xen-syms.symbols |grep do_domain_pause
0x82d080104920 t domain.c#do_domain_pause
$ nm --defined xen-syms|grep do_domain_paus
So that when we apply the patch again the .bss is cleared.
Otherwise we may find some variables containing old values.
The payloads may contain various .bss - especially if -fdata-sections
is used which can create .bss. sections.
Signed-off-by: Konrad Rzeszutek Wilk
---
Cc: Ross Lagerwall
Cc:
in case we want to patch at specific offsets inside
a function. (for example if we want to do NOP patching).
Signed-off-by: Konrad Rzeszutek Wilk
---
Cc: Konrad Rzeszutek Wilk
Cc: Ross Lagerwall
Cc: Stefano Stabellini
Cc: Julien Grall
Cc: Jan Beulich
Cc: Andrew Cooper
v3: First submission
It makes more sense for it to be there. However that
means the version.h has now a dependency on
as the Elf_Note is a macro.
The elfstructs.h has a dependency on types.h as well so
we need that. We cannot put that #include
in elfstructs.h as that file is used by tools and they
do not have such f
On Tue, 23 Aug 2016 19:33:06 +0200
"Luis R. Rodriguez" wrote:
> On Tue, Aug 23, 2016 at 11:26:33AM +1000, Nicholas Piggin wrote:
> > On Fri, 19 Aug 2016 14:34:02 -0700
> > mcg...@kernel.org wrote:
> > > +/**
> > > + * DOC: Standard ELF section use in Linux
> > > + *
> > > + * Linux makes use of
This run is configured for baseline tests only.
flight 67586 linux-3.18 real [real]
http://osstest.xs.citrite.net/~osstest/testlogs/logs/67586/
Regressions :-(
Tests which did not succeed and are blocking,
including tests which could not be run:
test-amd64-amd64-xl-qcow221 leak-check/check
On 24/08/16 00:51, Samuel Thibault wrote:
> Juergen Gross, on Tue 23 Aug 2016 17:16:06 +0200, wrote:
>> +printk("Xen Minimal OS (hvm)!\n");
>> +printk(" start_info: %p(VA)\n", si);
>> +printk(" shared_inf: %p(VA)\n", HYPERVISOR_shared_info);
>> +printk(" modlist: 0x%lx(PA)\n",
On 23/08/16 23:03, Samuel Thibault wrote:
> hello,
>
> Juergen Gross, on Tue 23 Aug 2016 17:15:52 +0200, wrote:
>> When running in HVMlite mode we need to setup the hypercall page by
>> ourself.
>>
>> Signed-off-by: Juergen Gross
>> ---
>>
>> +#else /* ifdef __x86_64__ */
>> +#error "Unsupporte
On 23/08/16 22:51, Samuel Thibault wrote:
> Hello,
>
> It looks good, there is just:
>
> Juergen Gross, on Tue 23 Aug 2016 17:15:51 +0200, wrote:
>> +.macro PTES prot, addr, incr, idx, end
>
> Please document what prot, addr, incr, idx, end represent precisely.
Okay.
>
>> +.if
This run is configured for baseline tests only.
flight 67587 ovmf real [real]
http://osstest.xs.citrite.net/~osstest/testlogs/logs/67587/
Perfect :-)
All tests in this flight passed as required
version targeted for testing:
ovmf 4962fcfa7d265824f01f74d782d5ed841ec8a72f
baseline v
95 matches
Mail list logo
