On Mon, Mar 17, 2025 at 05:07:22PM -0700, Stefano Stabellini wrote:
> On Mon, 17 Mar 2025, Roger Pau Monne wrote:
> > In preparation for using the XTF selftests to smoke test the FreeBSD based
> > Xen builds.
> >
> > Signed-off-by: Roger Pau Monné
> > ---
> > I've used the current version of Free
On 18/03/2025 5:35 pm, Roger Pau Monne wrote:
> mkelf32 attempt to check that the program header defined NOTE segment falls
> inside of the LOAD segment, as the build-id should be loaded for Xen at
> runtime to check.
>
> However the current code doesn't take into account the LOAD program header
>
As a result of relocations now being applied after the trampoline has been
copied into the low 1MB region, there's no need for a single .init section
that's writable, as .init.text is no longer modified.
Remove the bodge and fallback to the layout used by ELF images with an
.init.text and .init.da
On 18/03/2025 5:35 pm, Roger Pau Monne wrote:
> It's unclear why -N is being used in the first place. It was added by
> commit 40828c657dd0c back in 2004 without any justification.
>
> When building a PE image it's actually detrimental to forcefully set the
> .text section as writable. The GNU LD
On 17/03/2025 5:15 pm, Andrew Cooper wrote:
> On 17/03/2025 8:43 am, Jan Beulich wrote:
>> On 14.03.2025 18:59, Andrew Cooper wrote:
>>> This is an unnecessary indirection.
>>>
>>> Signed-off-by: Andrew Cooper
>> Reviewed-by: Jan Beulich
> So, this is the patch causing build chaos. I've reproduc
On 18/03/2025 5:35 pm, Roger Pau Monne wrote:
> As a result of relocations now being applied after the trampoline has been
> copied into the low 1MB region, there's no need for a single .init section
> that's writable, as .init.text is no longer modified.
>
> Remove the bodge and fallback to the la
On 18.03.2025 14:11, Andrew Cooper wrote:
> On 18/03/2025 9:19 am, Roger Pau Monne wrote:
>> The call to ioremap_wc() in video_init() will always fail, because
>> video_init() is called ahead of vm_init_type(), and so the underlying
>> __vmap() call will fail to allocate the linear address space.
>
On Tue, 18 Mar 2025, Jan Beulich wrote:
> On 18.03.2025 00:55, Stefano Stabellini wrote:
> > On Mon, 17 Mar 2025, Jason Andryuk wrote:
> >> On 2025-03-17 10:22, Jan Beulich wrote:
> >>> On 06.03.2025 23:03, Jason Andryuk wrote:
> --- a/xen/include/xsm/dummy.h
> +++ b/xen/include/xsm/dummy
When enabling UBSAN with clang, the following error is triggered during the
build:
common/wait.c:154:9: error: symbol '.L_wq_resume' is already defined
154 | "push %%rbx; push %%rbp; push %%r12;"
| ^
:1:121: note: instantiated into assembly here
1 | push %rbx; p
Introduce an additional Kconfig check to only offer the option if the
compiler supports -fsanitize=undefined.
We no longer use Travis CI, so the original motivation for not enabling
UBSAN might no longer present. Regardless, the option won't be present in
the first place if the compiler doesn't s
UBSAN complains with:
UBSAN: Undefined behaviour in arch/x86/mm/shadow/private.h:515:30
pointer operation overflowed 82e0 to 82dfffe0
[...]
Xen call trace:
[] R common/ubsan/ubsan.c#ubsan_epilogue+0xa/0xc0
[] F
lib/xxhash64.c#__ubsan_handle_pointer_overflow+0xcb/0x100
The call to ioremap_wc() in video_init() will always fail, because
video_init() is called ahead of vm_init_type(), and so the underlying
__vmap() call will fail to allocate the linear address space.
Fix by reverting to the previous behavior and using ioremap() for the VGA
text buffer.
Fixes: 81d1
Hello,
This started as a series to fix UBSAN when using clang, and the first
patch do fix that.
Finally patches 2 to 4 fix bugs in the code highlighted by the clang UB
pointer arithmetic detection.
Patch 5 was the original goal of the series: be able to enable UBSAN for
randconfig.
Thanks, Roge
On 18.03.2025 10:10, Mykyta Poturai wrote:
> On 15.01.24 11:35, Jan Beulich wrote:
>> On 14.01.2024 11:01, Mykyta Poturai wrote:
>>> --- a/xen/include/public/hvm/dm_op.h
>>> +++ b/xen/include/public/hvm/dm_op.h
>>> @@ -444,6 +444,17 @@ struct xen_dm_op_nr_vcpus {
>>> };
>>> typedef struct xen_d
On 18/03/2025 9:19 am, Roger Pau Monne wrote:
> Introduce an additional Kconfig check to only offer the option if the
> compiler supports -fsanitize=undefined.
>
> We no longer use Travis CI, so the original motivation for not enabling
> UBSAN might no longer present. Regardless, the option won't
On 3/17/25 9:07 PM, Luca Fancellu wrote:
From: Penny Zheng
ARM MPU system doesn't need to use paging memory pool, as MPU memory
mapping table at most takes only one 4KB page, which is enough to
manage the maximum 255 MPU memory regions, for all EL2 stage 1
translation and EL1 stage 2 translatio
On 18/03/2025 9:19 am, Roger Pau Monne wrote:
> The call to ioremap_wc() in video_init() will always fail, because
> video_init() is called ahead of vm_init_type(), and so the underlying
> __vmap() call will fail to allocate the linear address space.
>
> Fix by reverting to the previous behavior an
For the record, the rest of the series doesn't require this patch. I just
thought it was a strictly net-positive improvement on the current behaviour.
On Mon Mar 17, 2025 at 4:33 PM GMT, Jan Beulich wrote:
> On 14.03.2025 18:24, Alejandro Vallejo wrote:
> > As it is, it's incredibly easy for a bug
On Tue, Mar 18, 2025 at 09:29:45AM +0100, Roger Pau Monne wrote:
> Relax the limitation on MSI register writes, and only apply it when the
> system is in active state. For example AMD IOMMU drivers rely on using
> set_msi_affinity() to force an MSI register write on resume from
> suspension.
>
>
On Tue, Mar 18, 2025 at 09:36:37AM +0100, Jan Beulich wrote:
> On 18.03.2025 09:29, Roger Pau Monne wrote:
> > --- a/xen/drivers/passthrough/amd/iommu_intr.c
> > +++ b/xen/drivers/passthrough/amd/iommu_intr.c
> > @@ -546,7 +546,7 @@ int cf_check amd_iommu_msi_msg_update_ire(
> > rc = update_in
On Mon, Mar 17, 2025 at 05:09:25PM -0700, Stefano Stabellini wrote:
> On Mon, 17 Mar 2025, Roger Pau Monne wrote:
> > Introduce a basic set of smoke tests using the XTF selftest image, and run
> > them on QEMU. Use the matrix keyword to create a different task for each
> > XTF flavor on each FreeB
On 18.03.25 12:11, Jan Beulich wrote:
> On 18.03.2025 10:10, Mykyta Poturai wrote:
>> On 15.01.24 11:35, Jan Beulich wrote:
>>> On 14.01.2024 11:01, Mykyta Poturai wrote:
--- a/xen/include/public/hvm/dm_op.h
+++ b/xen/include/public/hvm/dm_op.h
@@ -444,6 +444,17 @@ struct xen_dm_op_n
This needs to be ahead of everything. Right now, it is after xen/init.h being
included for -DINIT_SECTIONS_ONLY
# 1 "./include/xen/compiler.h" 1
# 83 "./include/xen/compiler.h"
# 1 "./include/xen/init.h" 1
# 62 "./include/xen/init.h"
typedef int (*initcall_t)(void);
typedef void (*exi
On Tue, Mar 18, 2025 at 04:50:58PM +0100, Jan Beulich wrote:
> On 18.03.2025 16:35, Roger Pau Monné wrote:
> > On Tue, Mar 18, 2025 at 03:33:03PM +0100, Jan Beulich wrote:
> >> On 18.03.2025 10:19, Roger Pau Monne wrote:
> >>> --- a/xen/arch/x86/include/asm/x86_64/uaccess.h
> >>> +++ b/xen/arch/x86
On Tue, Mar 18, 2025 at 11:14:59AM +0100, Jan Beulich wrote:
> On 18.03.2025 09:54, Roger Pau Monné wrote:
> > On Tue, Mar 18, 2025 at 09:36:37AM +0100, Jan Beulich wrote:
> >> On 18.03.2025 09:29, Roger Pau Monne wrote:
> >>> --- a/xen/drivers/passthrough/amd/iommu_intr.c
> >>> +++ b/xen/drivers/p
Hi Michal,
> On 18 Mar 2025, at 09:00, Michal Orzel wrote:
>
> We are missing a way to detect whether a user provided a value for
> nr_spis equal to 0 or did not provide any value (default is also 0) which
> can cause issues when calculated nr_spis is > 0 and the value from domain
> config is 0.
On 18.03.2025 16:39, Nathan Studer wrote:
>
> On 17/03/25 05:31, Jan Beulich wrote:
>> Even before its recent movement to the scheduler's private data structure it
>> looks
>> to have been wrong to update the field under lock, but then read it with the
>> lock
>> no longer held.
>>
>> Coverity-I
Following on from 250d87dc3ff9 ("x86/msi: Change __msi_set_enable() to
take pci_sbdf_t"), struct amd_iommu has its seg and bdf fields
backwards with relation to pci_sbdf_t. Instead of regenerating sbdf_t
from seg+bdf, make the struct contain pci_sbdf_t directly, which simplifies
code.
I've avoided
On 18.03.2025 16:31, Roger Pau Monné wrote:
> On Tue, Mar 18, 2025 at 03:28:32PM +0100, Jan Beulich wrote:
>> On 18.03.2025 14:11, Andrew Cooper wrote:
>>> On 18/03/2025 9:19 am, Roger Pau Monne wrote:
The call to ioremap_wc() in video_init() will always fail, because
video_init() is call
On 18.03.2025 16:35, Roger Pau Monné wrote:
> On Tue, Mar 18, 2025 at 03:33:03PM +0100, Jan Beulich wrote:
>> On 18.03.2025 10:19, Roger Pau Monne wrote:
>>> --- a/xen/arch/x86/include/asm/x86_64/uaccess.h
>>> +++ b/xen/arch/x86/include/asm/x86_64/uaccess.h
>>> @@ -9,9 +9,9 @@
>>> * a secondary m
On 18.03.2025 17:47, Roger Pau Monné wrote:
> On Tue, Mar 18, 2025 at 04:50:58PM +0100, Jan Beulich wrote:
>> On 18.03.2025 16:35, Roger Pau Monné wrote:
>>> On Tue, Mar 18, 2025 at 03:33:03PM +0100, Jan Beulich wrote:
On 18.03.2025 10:19, Roger Pau Monne wrote:
> --- a/xen/arch/x86/includ
From: Frédéric Pierret (fepitre)
Ensure to have a realpath for XEN_ROOT else it fails to substitute
properly pathes in strings sections
Signed-off-by: Frédéric Pierret (fepitre)
[use cc-option-add]
Signed-off-by: Marek Marczykowski-Górecki
---
tools/Rules.mk| 2 ++
xen/Makefile
From: Frédéric Pierret (fepitre)
It makes the build reproducible with fileordering flags
Signed-off-by: Frédéric Pierret (fepitre)
---
docs/xen-headers | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/docs/xen-headers b/docs/xen-headers
index 8c434d77e20e..98ffe814500b 10075
Few patches from qubes patch queue that improve reproducible builds.
Frédéric Pierret (fepitre) (2):
docs/xen-headers: use alphabetical sorting for @incontents
Strip build path directories in tools, xen and xen/arch/x86
docs/xen-headers | 2 +-
tools/Rules.mk| 2 ++
xen/Makefile
On Mon, Mar 17, 2025 at 04:48:05PM -0700, Stefano Stabellini wrote:
> On Mon, 17 Mar 2025, Roger Pau Monne wrote:
> > Instead use env to find the location of expect.
> >
> > Additionally do not use the -f flag, as it's only meaningful when passing
> > arguments on the command line, which we never
On 18/03/2025 9:19 am, Roger Pau Monne wrote:
> When enabling UBSAN with clang, the following error is triggered during the
> build:
>
> common/wait.c:154:9: error: symbol '.L_wq_resume' is already defined
> 154 | "push %%rbx; push %%rbp; push %%r12;"
> | ^
> :1:121: note: i
On 18/03/2025 9:19 am, Roger Pau Monne wrote:
> UBSAN complains with:
>
> UBSAN: Undefined behaviour in common/compat/memory.c:90:9
> pointer operation overflowed 82008000 to 02008000
> [...]
> Xen call trace:
>[] R common/ubsan/ubsan.c#ubsan_epilogue+0xa/0xc0
>[] F
> lib/x
We are missing a way to detect whether a user provided a value for
nr_spis equal to 0 or did not provide any value (default is also 0) which
can cause issues when calculated nr_spis is > 0 and the value from domain
config is 0. Fix it by setting default value for nr_spis to UINT32_MAX
(max supporte
On 18.03.2025 00:55, Stefano Stabellini wrote:
> On Mon, 17 Mar 2025, Jason Andryuk wrote:
>> On 2025-03-17 10:22, Jan Beulich wrote:
>>> On 06.03.2025 23:03, Jason Andryuk wrote:
--- a/xen/include/xsm/dummy.h
+++ b/xen/include/xsm/dummy.h
@@ -95,7 +95,11 @@ static always_inline int
On 17.03.2025 20:20, Andrew Cooper wrote:
> Something I overlooked when last cleaning up exception handling is that a TSS
> is not necessary if IST isn't configured, and IST isn't necessary until we're
> running guest code.
>
> Introduce early_traps_init() which is far more minimal than init_idt_t
On Tue, Mar 18, 2025 at 03:36:45PM +0100, Jan Beulich wrote:
> On 18.03.2025 13:53, Andrew Cooper wrote:
> > On 18/03/2025 9:19 am, Roger Pau Monne wrote:
> >> UBSAN complains with:
> >>
> >> UBSAN: Undefined behaviour in arch/x86/mm/shadow/private.h:515:30
> >> pointer operation overflowed 82e
Following a similar change to amd_iommu struct, make two more structs
take pci_sbdf_t directly instead of seg and bdf separately. This lets us
drop several conversions from the latter to the former and simplifies
several comparisons and assignments.
Signed-off-by: Andrii Sultanov
---
This parti
On Tue, Mar 18, 2025 at 03:28:32PM +0100, Jan Beulich wrote:
> On 18.03.2025 14:11, Andrew Cooper wrote:
> > On 18/03/2025 9:19 am, Roger Pau Monne wrote:
> >> The call to ioremap_wc() in video_init() will always fail, because
> >> video_init() is called ahead of vm_init_type(), and so the underlyi
Relax the limitation on MSI register writes, and only apply it when the
system is in active state. For example AMD IOMMU drivers rely on using
set_msi_affinity() to force an MSI register write on resume from
suspension.
The original patch intention was to reduce the number of MSI register
writes
Step-by-step, use pci_sbdf_t directly where appropriate instead of
handling seg and bdf separately. This removes conversions, reduces code
size and simplifies code in general.
Andrii Sultanov (3):
drivers: Change amd_iommu struct to contain pci_sbdf_t, simplify code
drivers: Change find_iommu_
On Tue, Mar 18, 2025 at 03:33:03PM +0100, Jan Beulich wrote:
> On 18.03.2025 10:19, Roger Pau Monne wrote:
> > --- a/xen/arch/x86/include/asm/x86_64/uaccess.h
> > +++ b/xen/arch/x86/include/asm/x86_64/uaccess.h
> > @@ -9,9 +9,9 @@
> > * a secondary mapping installed, which needs to be used for su
On 17/03/25 05:31, Jan Beulich wrote:
> Even before its recent movement to the scheduler's private data structure it
> looks
> to have been wrong to update the field under lock, but then read it with the
> lock
> no longer held.
>
> Coverity-ID: 1644500
> Fixes: 9f0c658baedc ("arinc: add cpu-po
On 18.03.2025 09:29, Roger Pau Monne wrote:
> --- a/xen/drivers/passthrough/amd/iommu_intr.c
> +++ b/xen/drivers/passthrough/amd/iommu_intr.c
> @@ -546,7 +546,7 @@ int cf_check amd_iommu_msi_msg_update_ire(
> rc = update_intremap_entry_from_msi_msg(iommu, bdf, nr,
>
After 8e60d47cf011 writes from __setup_msi_irq() will no longer be
propagated to the MSI registers if the IOMMU IRTE was already allocated.
Given the purpose of __setup_msi_irq() is MSI initialization, always
propagate the write to the hardware, regardless of whether the IRTE was
already allocated.
On 14.03.25 22:37, Nico Pache wrote:
Update the NR_BALLOON_PAGES counter when pages are added to or
removed from the Xen balloon.
Signed-off-by: Nico Pache
Reviewed-by: Juergen Gross
Juergen
OpenPGP_0xB0DE9DD628BF132F.asc
Description: OpenPGP public key
OpenPGP_signature.asc
Descriptio
On 17/03/2025 21:07, Luca Fancellu wrote:
>
>
> From: Penny Zheng
>
> ARM MPU system doesn't need to use paging memory pool, as MPU memory
> mapping table at most takes only one 4KB page, which is enough to
> manage the maximum 255 MPU memory regions, for all EL2 stage 1
> translation and EL
On 18.03.2025 14:39, Andrew Cooper wrote:
> This needs to be ahead of everything. Right now, it is after xen/init.h being
> included for -DINIT_SECTIONS_ONLY
>
> # 1 "./include/xen/compiler.h" 1
> # 83 "./include/xen/compiler.h"
> # 1 "./include/xen/init.h" 1
> # 62 "./include/xen/init.h"
On 18.03.2025 14:31, Mykyta Poturai wrote:
> On 18.03.25 12:11, Jan Beulich wrote:
>> On 18.03.2025 10:10, Mykyta Poturai wrote:
>>> On 15.01.24 11:35, Jan Beulich wrote:
On 14.01.2024 11:01, Mykyta Poturai wrote:
> --- a/xen/include/public/hvm/dm_op.h
> +++ b/xen/include/public/hvm/dm
On 18.03.2025 10:19, Roger Pau Monne wrote:
> --- a/xen/arch/x86/include/asm/x86_64/uaccess.h
> +++ b/xen/arch/x86/include/asm/x86_64/uaccess.h
> @@ -9,9 +9,9 @@
> * a secondary mapping installed, which needs to be used for such accesses in
> * the PV case, and will also be used for HVM to avoi
On 18/03/2025 9:19 am, Roger Pau Monne wrote:
> UBSAN complains with:
>
> UBSAN: Undefined behaviour in arch/x86/mm/shadow/private.h:515:30
> pointer operation overflowed 82e0 to 82dfffe0
> [...]
> Xen call trace:
>[] R common/ubsan/ubsan.c#ubsan_epilogue+0xa/0xc0
>[] F
On 18.03.2025 13:53, Andrew Cooper wrote:
> On 18/03/2025 9:19 am, Roger Pau Monne wrote:
>> UBSAN complains with:
>>
>> UBSAN: Undefined behaviour in arch/x86/mm/shadow/private.h:515:30
>> pointer operation overflowed 82e0 to 82dfffe0
>> [...]
>> Xen call trace:
>>[] R comm
On 18/03/25 03:42, Juergen Gross wrote:
> On 18.03.25 08:34, Anderson Choi wrote:
> > xen panic is observed with the following configuration.
> >
> > 1. Debug xen build (CONFIG_DEBUG=y)
> > 2. dom1 of an ARINC653 domain
> > 3. shutdown dom1 with xl command
> >
> > $ xl shutdown
> >
> > (XEN) *
On Tue, Mar 18, 2025 at 07:24:31AM +0100, Jan Beulich wrote:
> On 17.03.2025 18:38, Roger Pau Monné wrote:
> > On Mon, Mar 17, 2025 at 05:11:56PM +0100, Jan Beulich wrote:
> >> On 17.03.2025 16:56, Roger Pau Monné wrote:
> >>> On Sat, Mar 15, 2025 at 12:02:50AM +, Andrew Cooper wrote:
> On
On Mon, Mar 17, 2025 at 06:54:06PM +0100, Marek Marczykowski-Górecki wrote:
> On Mon, Mar 17, 2025 at 06:35:08PM +0100, Roger Pau Monné wrote:
> > On Mon, Mar 17, 2025 at 05:35:51PM +0100, Marek Marczykowski-Górecki wrote:
> > > On Mon, Mar 17, 2025 at 04:56:15PM +0100, Roger Pau Monné wrote:
> > >
On 18.03.25 08:34, Anderson Choi wrote:
xen panic is observed with the following configuration.
1. Debug xen build (CONFIG_DEBUG=y)
2. dom1 of an ARINC653 domain
3. shutdown dom1 with xl command
$ xl shutdown
(XEN)
(XEN) Panic on CPU 2:
(XEN) Assertion
Jan,
> EXT email: be mindful of links/attachments.
>
> On 18.03.2025 05:00, Anderson Choi wrote:
>> xen panic is observed with the following configuration.
>>
>> 1. Debug xen build (CONFIG_DEBUG=y)
>> 2. dom1 of an ARINC653 domain
>> 3. shutdown dom1 with xl command
>>
>> $ xl shutdown
>>
>>
On 17.03.2025 20:20, Andrew Cooper wrote:
> cpu_init() doesn't particularly belong in trap_init(). This brings the BSP
> more in line with the APs.
>
> No functional change.
>
> Signed-off-by: Andrew Cooper
Acked-by: Jan Beulich
Clarify that trampoline_setup is only used for EFI when booted using the
multiboot2 entry point.
Signed-off-by: Roger Pau Monné
---
xen/arch/x86/boot/head.S | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/xen/arch/x86/boot/head.S b/xen/arch/x86/boot/head.S
index 1b3bd16fe57
mkelf32 attempt to check that the program header defined NOTE segment falls
inside of the LOAD segment, as the build-id should be loaded for Xen at
runtime to check.
However the current code doesn't take into account the LOAD program header
segment offset when calculating overlap with the NOTE seg
On 18/03/2025 5:35 pm, Roger Pau Monne wrote:
> Relocations are now applied after having moved the trampoline, so there's no
> reason to warn about relocations to read-only sections. The logic that
> apply the relocations would make sure they are applied against writable
> mappings.
>
> Signed-off
It's unclear why -N is being used in the first place. It was added by
commit 40828c657dd0c back in 2004 without any justification.
When building a PE image it's actually detrimental to forcefully set the
.text section as writable. The GNU LD man page contains the following
warning regarding the
On 18/03/2025 7:05 pm, Frediano Ziglio wrote:
> On Tue, Mar 18, 2025 at 5:36 PM Roger Pau Monne wrote:
>> Change the order relocations are applied. Currently the trampoline is
>> patched for relocations before being copied to the low 1MB region. Change
>> the order and instead copy the trampolin
Rename max_init_domid to domid_top to align with its usage in the code
(Arm), where it represents the upper boundary of the non-system domain
ID range.
Relocate the domid_top declaration to an architecture-independent
location.
Signed-off-by: Denis Mukhin
---
xen/arch/arm/dom0less-build.c
The little endian implementation of bitmap_to_xenctl_bitmap leads to
unnecessary xmallocs and xfrees. Given that Xen only supports little
endian architectures, it is worth optimizing.
This patch removes the need for the xmalloc on little endian
architectures.
Signed-off-by: Stefano Stabellini
d
Add console_get_focus() as a console public API to the retrieve current
console owner domain ID.
Make console_{get,put}_domain() private and simplify vpl011 code a bit.
Signed-off-by: Denis Mukhin
---
xen/arch/arm/vpl011.c | 5 +
xen/drivers/char/console.c | 9 +++--
xen/include/xe
UBSAN complains with:
UBSAN: Undefined behaviour in common/compat/memory.c:90:9
pointer operation overflowed 82008000 to 02008000
[...]
Xen call trace:
[] R common/ubsan/ubsan.c#ubsan_epilogue+0xa/0xc0
[] F
lib/xxhash64.c#__ubsan_handle_pointer_overflow+0xcb/0x100
[] F li
On 18/03/2025 5:35 pm, Roger Pau Monne wrote:
> Clarify that trampoline_setup is only used for EFI when booted using the
> multiboot2 entry point.
>
> Signed-off-by: Roger Pau Monné
> ---
> xen/arch/x86/boot/head.S | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/xen/arc
On 18/03/2025 5:35 pm, Roger Pau Monne wrote:
> After having removed the -N option from the linker script invocation, and
> also having removed the merging of the .init.text and .init.data sections
> on x86, there should be no remaining RWX load segments. Do not silence the
> GNU LD warning.
>
> S
Relocations are now applied after having moved the trampoline, so there's no
reason to warn about relocations to read-only sections. The logic that
apply the relocations would make sure they are applied against writable
mappings.
Signed-off-by: Roger Pau Monné
---
xen/arch/x86/efi/mkreloc.c | 5
Change the order relocations are applied. Currently the trampoline is
patched for relocations before being copied to the low 1MB region. Change
the order and instead copy the trampoline first to the low 1MB region and
then apply the relocations.
This will allow making .init.text section read-onl
Hello,
The aim of the series is to generate a Xen image with no write and
execute sections, so that the PE binary can be NX_COMPAT.
The main change for achieving this is changing the order in which the
trampoline relocation are applied. To avoid having write-execute
sections apply the trampoline
On Tue, Mar 18, 2025 at 5:36 PM Roger Pau Monne wrote:
>
> Change the order relocations are applied. Currently the trampoline is
> patched for relocations before being copied to the low 1MB region. Change
> the order and instead copy the trampoline first to the low 1MB region and
> then apply th
On Monday, March 17th, 2025 at 1:30 AM, Jan Beulich wrote:
>
>
> On 15.03.2025 02:00, dm...@proton.me wrote:
>
> > Add new symbol APIC_VECTOR_VALID to replace open-coded value 16 in
> > LAPIC and virtual LAPIC code.
>
>
> First a good name is needed to make such a change. APIC_VECTOR_VALID
> coul
Update the symbol name in preparation for the semantic change
to the physical console input owner domain identifier.
No functional change.
Signed-off-by: Denis Mukhin
---
xen/drivers/char/console.c | 23 ++-
1 file changed, 10 insertions(+), 13 deletions(-)
diff --git a/xen
Add new flag in domain structure for marking permission to intercept
the physical console input by the domain.
Update console input switch logic accordingly.
Signed-off-by: Denis Mukhin
---
xen/arch/arm/vpl011.c | 2 ++
xen/arch/x86/pv/shim.c | 2 ++
xen/common/domain.c| 2 +
Move domain ID allocation to a dedicated function domid_alloc() and
use it during domain creation.
Update domid_top within domid_alloc() to reflect the highest known
domain ID.
Initialize domid_top using the result of get_initial_domain_id().
Allocation algorithm:
- If an explicit domain ID is p
Update the name to emphasize the physical console input switch to a
new owner domain following the naming notation in the console driver.
No functional change.
Signed-off-by: Denis Mukhin
---
xen/drivers/char/console.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/xe
Switch console_focus address space from integers mapped to domain IDs to
direct domain IDs, simplifying the console input switching code.
Introduce console_set_focus() to set the console owner domain identifier.
Signed-off-by: Denis Mukhin
---
xen/drivers/char/console.c | 81 ---
Remove trailing whitespaces in the console driver.
No functional change.
Signed-off-by: Denis Mukhin
---
xen/drivers/char/console.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/xen/drivers/char/console.c b/xen/drivers/char/console.c
index ba428199d2..c3150fbdb7 1006
Evtchn fifos are not needed on smaller systems; the older interface is
lightweight and sufficient. Make it possible to disable evtchn fifo.
Signed-off-by: Stefano Stabellini
diff --git a/xen/common/Kconfig b/xen/common/Kconfig
index a6aa2c5c14..14d82923c5 100644
--- a/xen/common/Kconfig
+++ b/xe
Currently, on x86, console input can be rotated in round-robin manner
only between dom0, PV shim, and Xen itself. On Arm the input rotation
can include domUs with vpl011.
The patch series introduces the concept of "console focus", which is
defined as the ID of the domain that currently owns the
After having removed the -N option from the linker script invocation, and
also having removed the merging of the .init.text and .init.data sections
on x86, there should be no remaining RWX load segments. Do not silence the
GNU LD warning.
Signed-off-by: Roger Pau Monné
---
xen/Makefile | 2 --
On Tue, Mar 18, 2025 at 06:01:46PM +0100, Jan Beulich wrote:
> On 18.03.2025 17:47, Roger Pau Monné wrote:
> > On Tue, Mar 18, 2025 at 04:50:58PM +0100, Jan Beulich wrote:
> >> On 18.03.2025 16:35, Roger Pau Monné wrote:
> >>> On Tue, Mar 18, 2025 at 03:33:03PM +0100, Jan Beulich wrote:
> On 1
The LRET is detached from the PUSHes which set it up, and this is about to get
worse with the changes to trampoline relocation. For the sake of one variable
read, the complexity is not worth it.
Reorder the logic to copy the trampoline into place, then switch stack and
enter the trampoline.
No f
On 18/03/2025 5:35 pm, Roger Pau Monne wrote:
> diff --git a/xen/arch/x86/boot/reloc-trampoline.c
> b/xen/arch/x86/boot/reloc-trampoline.c
> index e35e7c78aa86..ac54aef14eaf 100644
> --- a/xen/arch/x86/boot/reloc-trampoline.c
> +++ b/xen/arch/x86/boot/reloc-trampoline.c
> @@ -20,19 +20,19 @@ void
On 18.03.2025 09:54, Roger Pau Monné wrote:
> On Tue, Mar 18, 2025 at 09:36:37AM +0100, Jan Beulich wrote:
>> On 18.03.2025 09:29, Roger Pau Monne wrote:
>>> --- a/xen/drivers/passthrough/amd/iommu_intr.c
>>> +++ b/xen/drivers/passthrough/amd/iommu_intr.c
>>> @@ -546,7 +546,7 @@ int cf_check amd_io
On Tue, Mar 4, 2025 at 1:23 PM Frediano Ziglio
wrote:
>
> Describe the usage of devices 5853:0002 and 5853:C000.
>
> Signed-off-by: Frediano Ziglio
> ---
> Changes since v1:
> - address minor comments
> ---
> docs/man/xen-pci-device-reservations.7.pod | 7 +++
> 1 file changed, 7 insertions(
92 matches
Mail list logo