Xen 4.19 release status tracking list [ May ]

Xen 4.19 release status tracking list [ May ]

[Oleksii]

Hello everyone,

I would like to share with you a list for status tracking based on Xen
ML and community members comments:

*** Arm ***:
  * Arm cache coloring [
https://lore.kernel.org/xen-devel/20240502165533.319988-1-carlo.nonato@minervasys.tech/
]:
    - new patch series version [v8] was sent

  * PCI devices passthrough on Arm, part 3 [
https://lore.kernel.org/xen-devel/20240202213321.1920347-1-stewart.hildebrand@amd.com/
]
 
  * DOMCTL-based guest magic region allocation for 11
domUs [
https://lore.kernel.org/xen-devel/20240409045357.236802-1-xin.wang2@amd.com/
]
     - new patch series verstion [v4] was sent
 
  * [XEN v6 0/3] xen/arm: Add emulation of Debug Data Transfer
Registers [
https://patchew.org/Xen/20240307123943.1991755-1-ayan.kumar.halder@amd.com/
]

*** PPC ***:
  * [PATCH v4 0/6] Early Boot Allocation on Power [
https://lore.kernel.org/xen-devel/cover.1712893887.git.sanastasio@raptorengineering.com/
]:
    new patch series version [v4] was sent

*** RISC-V ***:
  * [PATCH v8 00/17] Enable build of full Xen for RISC-V
https://lore.kernel.org/xen-devel/cover.1713347222.git.oleksii.kurochko@gmail.com/
]:
    - several patches were merged
    - new patch series version [v8] were sent


*** x86 ***:
  * [PATCH 0/4] iommu/x86: fixes/improvements for unity range checks [
https://lore.kernel.org/xen-devel/20240201170159.66330-1-roger.pau@citrix.com/
]:
    - almost patch series have been merged already except the patch:
        [PATCH 4/4] iommu/x86: make unity range checking more strict

  * [PATCH 0/8] x86: support AVX10.1 [
https://lore.kernel.org/xen-devel/298db76f-d0ee-4d47-931f-1baa1a7546cf@suse.com/
]:
    - two patches of patch series are waitng to merged/reviewed:
      [PATCH 1/4] amd-vi: fix IVMD memory type checks
      [PATCH 4/4] iommu/x86: make unity range checking more strict 
  
  * APX support?

  * [PATCH v4 0/8] x86emul: misc additions [
https://lore.kernel.org/xen-devel/9dd23064-c79e-4a50-9c71-c0e73b189944@suse.com/
]
   
  * [PATCH v2 00/12] VT-d: SATC handling; ATS: tidying[
https://lore.kernel.org/xen-devel/64b028be-2197-4951-ae5b-32f9eabfa84a@suse.com/
]:
    new version was sent [ v2 ]

  * [XEN PATCH 0/9] x86: parallelize AP bring-up during boot [
https://lore.kernel.org/xen-devel/cover.1699982111.git.krystian.hebel@3mdeb.com/
]

  * [PATCH v2 00/12] x86: memcpy() / memset() (non-)ERMS flavors plus
fallout [
https://lore.kernel.org/xen-devel/8f56a8f4-0482-932f-96a9-c791bebb4610@suse.com/
]
    - 6/12 are merged.
 
  * [PATCH v6 0/4] x86/pvh: Support relocating dom0 kernel [
https://patchew.org/Xen/20240327215102.136001-1-jason.andryuk@amd.com/
]
 
  * x86/spec-ctrl: IBPB improvements [
https://patchew.org/Xen/06591b64-2f05-a4cc-a2f3-a74c3c4a76d6@suse.com/
]


*** common ***:
  * annotate entry points with type and size" series:
    The bulk of this has gone in, but there'll want to be follow-ups.

  * [PATCH v2 (resend) 00/27] Remove the directmap [
https://lore.kernel.org/xen-devel/20240116192611.41112-1-eliasely@amazon.com/
]
    - 7/27 were merged.
 
  * [PATCH] move __read_mostly to xen/cache.h [
https://lore.kernel.org/xen-devel/f25eb5c9-7c14-6e23-8535-2c66772b333e@suse.com/
]

  * [XEN PATCH v2 1/3] xen: introduce STATIC_ASSERT_UNREACHABLE() [
https://lore.kernel.org/xen-devel/42fc6ae8d3eb802429d29c774502ff232340dc84.1706259490.git.federico.serafini@bugseng.com/
]

  * MISRA rules updates:
   - [PATCH v2] docs/misra/rules.rst update [
https://lore.kernel.org/xen-devel/alpine.DEB.2.22.394.2402131431070.1925432@ubuntu-linux-20-04-desktop/T/#maded3df1bebe68d0fe53c73e89f996ec395a39e5
]: 2/3 were merged.

   - [XEN PATCH v3 0/7] address violations of MISRA C Rule 20.7[
https://patchew.org/Xen/cover.1711700095.git.nicola.vetrini@bugseng.com/
]: new patch series version (v3) were sent.

   - [XEN PATCH v3 00/16] xen: address violation of MISRA C:2012
Directive 4.10 [
https://patchew.org/Xen/cover.1710145041.git.simone.ballarin@bugseng.com/
]: 2/16 were merged.

  * [PATCH v6 00/8] xen/spinlock: make recursive spinlocks a dedicated
type [ https://patchew.org/Xen/20240327152229.25847-1-jgross@suse.com/
]:
    - only one patch required to be merged:
     
https://patchew.org/Xen/20240327152229.25847-1-jgross@suse.com/20240327152229.25847-9-jgross@suse.com/
   
  * [PATCH 0/7] GRUB: Supporting Secure Boot of xen.gz [
https://patchew.org/Xen/20240313150748.791236-1-ross.lagerwall@citrix.com/
]:
   
  * [PATCH v6 0/7] MSI-X support with qemu in stubdomain, and other
related changes:
    - new patch version was sent
 
  * [PATCH 0/7] xen/bitops: Reduce the mess, starting with ffs() [
https://patchew.org/Xen/20240313172716.2325427-1-andrew.cooper3@citrix.com/
]:
 
  * [PATCH 0/7] xen/trace: Treewide API cleanup [
https://patchew.org/Xen/20240318163552.3808695-1-andrew.cooper3@citrix.com/
]:
    patches were merged to staging:
     - [PATCH 3/7] xen/rt: Clean up trace handling
     - [PATCH 4/7] xen/sched: Clean up trace handling
 
  * [PATCH v3 0/4] xenwatchdogd bugfixes and enhancements [
https://lore.kernel.org/xen-devel/20240411182023.56309-1-leigh@solinno.co.uk/
]:
    new patch series were sent.

  * [RFC XEN PATCH v7 0/5] Support device passthrough when dom0 is PVH
on Xen [
https://patchew.org/Xen/20240419035340.608833-1-Jiqian.Chen@amd.com/ ]:
    - new patch series were sent [v7].


*** Completed ***:
  *** Arm ***:
     * xen/arm64: Rework the MMU-off code (idmap) so it is self-
contained

  *** x86 ***:
     * tools: enable xenstore-stubdom to use 9pfs

  *** common ***:
     * NUMA: no need for asm/numa.h when !NUMA
     * xen: move BUG_ON(), WARN_ON(), ASSERT(), ASSERT_UNREACHABLE() to
xen/bug.h
     * xen/lib: introduce generic find next bit operations
     * Introduce generic headers
     * xen/livepatch: fixes for the pre-apply / post-revert hooks
     * limit passing around of cpu_user_regs

Please reply with items you would like to see in 4.19 so that people
know what is happening and prioritize accordingly.
You're welcome to provide a description and use cases of the feature
you're working on.

Have a nice week!

Best regards,
 Oleksii

Re: Xen 4.19 release status tracking list [ May ]

[Jan Beulich]

On 03.05.2024 18:54, Oleksii wrote:
> *** x86 ***:
>   * [PATCH 0/4] iommu/x86: fixes/improvements for unity range checks [
> https://lore.kernel.org/xen-devel/20240201170159.66330-1-roger.pau@citrix.com/
> ]:
>     - almost patch series have been merged already except the patch:
>         [PATCH 4/4] iommu/x86: make unity range checking more strict
> 
>   * [PATCH 0/8] x86: support AVX10.1 [
> https://lore.kernel.org/xen-devel/298db76f-d0ee-4d47-931f-1baa1a7546cf@suse.com/
> ]:
>     - two patches of patch series are waitng to merged/reviewed:
>       [PATCH 1/4] amd-vi: fix IVMD memory type checks
>       [PATCH 4/4] iommu/x86: make unity range checking more strict 

This sub-item doesn't really fit the topic, and seems to rather belong
to the earlier one? Where (correctly) 1/4 isn't listed anymore, for
having gone in. As to 4/4 - Roger, I'm not sure I can conclude what the
plan here was: Are you meaning to submit an updated version, or did we
rather settle on not further pursuing this?

>   * APX support?

I think you should drop this now. I'm throttling further work on the insn
emulator, as long as I have so many other patches there pending review.

>   * [PATCH v6 0/4] x86/pvh: Support relocating dom0 kernel [
> https://patchew.org/Xen/20240327215102.136001-1-jason.andryuk@amd.com/
> ]

There had been a v7, and all of that plus a follow-on adjustment have gone
in.

> *** common ***:
>   * annotate entry points with type and size" series:
>     The bulk of this has gone in, but there'll want to be follow-ups.

I think these follow-ups now want tracking on a per-arch basis (x86 and
Arm, with PPC and RISC-V filly done).

>   * [PATCH v2 (resend) 00/27] Remove the directmap [
> https://lore.kernel.org/xen-devel/20240116192611.41112-1-eliasely@amazon.com/
> ]
>     - 7/27 were merged.

Hmm, no, I don't think that one was. Other were, yes.

>   * [PATCH v6 0/7] MSI-X support with qemu in stubdomain, and other
> related changes:
>     - new patch version was sent

1/7 was committed.

>   * [PATCH v3 0/4] xenwatchdogd bugfixes and enhancements [
> https://lore.kernel.org/xen-devel/20240411182023.56309-1-leigh@solinno.co.uk/
> ]:
>     new patch series were sent.

Was there anything left from that series?

Jan

Re: Xen 4.19 release status tracking list [ May ]

[Oleksii]

On Mon, 2024-05-06 at 09:11 +0200, Jan Beulich wrote:
> On 03.05.2024 18:54, Oleksii wrote:
> > *** x86 ***:
> >   * [PATCH 0/4] iommu/x86: fixes/improvements for unity range
> > checks [
> > https://lore.kernel.org/xen-devel/20240201170159.66330-1-roger.pau@citrix.com/
> > ]:
> >     - almost patch series have been merged already except the
> > patch:
> >         [PATCH 4/4] iommu/x86: make unity range checking more
> > strict
> > 
> >   * [PATCH 0/8] x86: support AVX10.1 [
> > https://lore.kernel.org/xen-devel/298db76f-d0ee-4d47-931f-1baa1a7546cf@suse.com/
> > ]:
> >     - two patches of patch series are waitng to merged/reviewed:
> >       [PATCH 1/4] amd-vi: fix IVMD memory type checks
> >       [PATCH 4/4] iommu/x86: make unity range checking more strict 
> 
> This sub-item doesn't really fit the topic, and seems to rather
> belong
> to the earlier one? 
Thanks for correction, it is belonged to previous topic.

> Where (correctly) 1/4 isn't listed anymore, for
> having gone in. As to 4/4 - Roger, I'm not sure I can conclude what
> the
> plan here was: Are you meaning to submit an updated version, or did
> we
> rather settle on not further pursuing this?
> 
> >   * APX support?
> 
> I think you should drop this now. I'm throttling further work on the
> insn
> emulator, as long as I have so many other patches there pending
> review.
> 
> >   * [PATCH v6 0/4] x86/pvh: Support relocating dom0 kernel [
> > https://patchew.org/Xen/20240327215102.136001-1-jason.andryuk@amd.com/
> > ]
> 
> There had been a v7, and all of that plus a follow-on adjustment have
> gone
> in.
> 
> > *** common ***:
> >   * annotate entry points with type and size" series:
> >     The bulk of this has gone in, but there'll want to be follow-
> > ups.
> 
> I think these follow-ups now want tracking on a per-arch basis (x86
> and
> Arm, with PPC and RISC-V filly done).
> 
> >   * [PATCH v2 (resend) 00/27] Remove the directmap [
> > https://lore.kernel.org/xen-devel/20240116192611.41112-1-eliasely@amazon.com/
> > ]
> >     - 7/27 were merged.
> 
> Hmm, no, I don't think that one was. Other were, yes.
> 
> >   * [PATCH v6 0/7] MSI-X support with qemu in stubdomain, and other
> > related changes:
> >     - new patch version was sent
> 
> 1/7 was committed.
> 
> >   * [PATCH v3 0/4] xenwatchdogd bugfixes and enhancements [
> > https://lore.kernel.org/xen-devel/20240411182023.56309-1-leigh@solinno.co.uk/
> > ]:
> >     new patch series were sent.
> 
> Was there anything left from that series?
I double-checked and everything has been merged. Thanks for correction.

I'm considering writing a script to automate the process of checking
whether the patches in a patch series have been merged or not.
Currently, I'm doing it manually, so there's a chance I might miss
something. It would be helpful if such a script already exists.

~ Oleksii

Re: Xen 4.19 release status tracking list [ May ]

[Roger Pau Monné]

On Fri, May 03, 2024 at 06:54:40PM +0200, Oleksii wrote:
> Hello everyone,
> 
> I would like to share with you a list for status tracking based on Xen
> ML and community members comments:
> 
> *** Arm ***:
>   * Arm cache coloring [
> https://lore.kernel.org/xen-devel/20240502165533.319988-1-carlo.nonato@minervasys.tech/
> ]:
>     - new patch series version [v8] was sent
> 
>   * PCI devices passthrough on Arm, part 3 [
> https://lore.kernel.org/xen-devel/20240202213321.1920347-1-stewart.hildebrand@amd.com/
> ]
>  
>   * DOMCTL-based guest magic region allocation for 11
> domUs [
> https://lore.kernel.org/xen-devel/20240409045357.236802-1-xin.wang2@amd.com/
> ]
>      - new patch series verstion [v4] was sent
>  
>   * [XEN v6 0/3] xen/arm: Add emulation of Debug Data Transfer
> Registers [
> https://patchew.org/Xen/20240307123943.1991755-1-ayan.kumar.halder@amd.com/
> ]
> 
> *** PPC ***:
>   * [PATCH v4 0/6] Early Boot Allocation on Power [
> https://lore.kernel.org/xen-devel/cover.1712893887.git.sanastasio@raptorengineering.com/
> ]:
>     new patch series version [v4] was sent
> 
> *** RISC-V ***:
>   * [PATCH v8 00/17] Enable build of full Xen for RISC-V
> https://lore.kernel.org/xen-devel/cover.1713347222.git.oleksii.kurochko@gmail.com/
> ]:
>     - several patches were merged
>     - new patch series version [v8] were sent
> 
> 
> *** x86 ***:
>   * [PATCH 0/4] iommu/x86: fixes/improvements for unity range checks [
> https://lore.kernel.org/xen-devel/20240201170159.66330-1-roger.pau@citrix.com/
> ]:
>     - almost patch series have been merged already except the patch:
>         [PATCH 4/4] iommu/x86: make unity range checking more strict
> 
>   * [PATCH 0/8] x86: support AVX10.1 [
> https://lore.kernel.org/xen-devel/298db76f-d0ee-4d47-931f-1baa1a7546cf@suse.com/
> ]:
>     - two patches of patch series are waitng to merged/reviewed:
>       [PATCH 1/4] amd-vi: fix IVMD memory type checks
>       [PATCH 4/4] iommu/x86: make unity range checking more strict 
>   
>   * APX support?
> 
>   * [PATCH v4 0/8] x86emul: misc additions [
> https://lore.kernel.org/xen-devel/9dd23064-c79e-4a50-9c71-c0e73b189944@suse.com/
> ]
>    
>   * [PATCH v2 00/12] VT-d: SATC handling; ATS: tidying[
> https://lore.kernel.org/xen-devel/64b028be-2197-4951-ae5b-32f9eabfa84a@suse.com/
> ]:
>     new version was sent [ v2 ]
> 
>   * [XEN PATCH 0/9] x86: parallelize AP bring-up during boot [
> https://lore.kernel.org/xen-devel/cover.1699982111.git.krystian.hebel@3mdeb.com/
> ]
> 
>   * [PATCH v2 00/12] x86: memcpy() / memset() (non-)ERMS flavors plus
> fallout [
> https://lore.kernel.org/xen-devel/8f56a8f4-0482-932f-96a9-c791bebb4610@suse.com/
> ]
>     - 6/12 are merged.
>  
>   * [PATCH v6 0/4] x86/pvh: Support relocating dom0 kernel [
> https://patchew.org/Xen/20240327215102.136001-1-jason.andryuk@amd.com/
> ]

Isn't this done? (at least the Xen side):

dfc9fab00378 x86/PVH: Support relocatable dom0 kernels

I have some minor stuff that was posted one day after feature
freeze:

xen/x86: support foreign mappings for HVM
https://lore.kernel.org/xen-devel/20240430165845.81696-1-roger.pau@citrix.com/

and

tools/xen-cpuid: switch to use cpu-policy defined names
https://lore.kernel.org/xen-devel/20240502114922.94288-1-roger.pau@citrix.com/

Which I think would be nice to have for the release.

The first one because it enables work on PVH-based QEMU stubdomains,
the second it's a nice cleanup that makes it easier to introduce new
CPUID features.

Some series that I posted before the freeze:

livepatch: minor bug fixes and improvements
https://lore.kernel.org/xen-devel/20240424081957.34326-1-roger.pau@citrix.com/
Patches 3 and 4 are pending review.

xen: introduce Kconfig function alignment option
https://lore.kernel.org/xen-devel/20240207145547.89689-1-roger.pau@citrix.com/
Patch 2 is missing review.

And then some patches that I don't expect to make progress:

x86/shutdown: change default reboot method preference
https://lore.kernel.org/xen-devel/20230915074347.94712-1-roger.pau@citrix.com/

x86/time: prefer CMOS over EFI_GET_TIME
https://lore.kernel.org/xen-devel/20240315114242.33309-1-roger.pau@citrix.com/

I find it quite funny how everyone complains about all downstreams
carrying similar patches to deal with firmware bugs, yet when I
attempt to formally post them I get zero interest from such
downstreams.  Maybe raising them here will help get some eyes.

Thanks, Roger.

Re: Xen 4.19 release status tracking list [ May ]

[Jan Beulich]

On 06.05.2024 11:26, Roger Pau Monné wrote:
> And then some patches that I don't expect to make progress:
> 
> x86/shutdown: change default reboot method preference
> https://lore.kernel.org/xen-devel/20230915074347.94712-1-roger.pau@citrix.com/
> 
> x86/time: prefer CMOS over EFI_GET_TIME
> https://lore.kernel.org/xen-devel/20240315114242.33309-1-roger.pau@citrix.com/
> 
> I find it quite funny how everyone complains about all downstreams
> carrying similar patches to deal with firmware bugs, yet when I
> attempt to formally post them I get zero interest from such
> downstreams.  Maybe raising them here will help get some eyes.

Well, as indicated on the Community Call I've meanwhile proposed a design
session (which won't really be about design) for Lisbon, as to maintenance
of our EFI interfacing. I think I had provided feedback on the shutdown
patch, and I didn't think there was much of a point repeating almost the
same for the time one.

As a minor remark on what you say above: It's clearly not "all"
downstreams, as we don't carry anything similar to either of the two
patches. But I'm pretty sure you meant "most" or "many" anyway.

Jan

Re: Xen 4.19 release status tracking list [ May ]

[Roger Pau Monné]

On Mon, May 06, 2024 at 11:35:00AM +0200, Jan Beulich wrote:
> On 06.05.2024 11:26, Roger Pau Monné wrote:
> > And then some patches that I don't expect to make progress:
> > 
> > x86/shutdown: change default reboot method preference
> > https://lore.kernel.org/xen-devel/20230915074347.94712-1-roger.pau@citrix.com/
> > 
> > x86/time: prefer CMOS over EFI_GET_TIME
> > https://lore.kernel.org/xen-devel/20240315114242.33309-1-roger.pau@citrix.com/
> > 
> > I find it quite funny how everyone complains about all downstreams
> > carrying similar patches to deal with firmware bugs, yet when I
> > attempt to formally post them I get zero interest from such
> > downstreams.  Maybe raising them here will help get some eyes.
> 
> Well, as indicated on the Community Call I've meanwhile proposed a design
> session (which won't really be about design) for Lisbon, as to maintenance
> of our EFI interfacing. I think I had provided feedback on the shutdown
> patch, and I didn't think there was much of a point repeating almost the
> same for the time one.

I know you are likely not OK with those, so I wasn't expecting input
from you.  We had discussions in the past where people complained that
our defaults are not OK because they lead to errors and panics on
production systems, and I'm surprised to find that none seem to have
cared about the above patches that attempt to address at least part of
this.

> As a minor remark on what you say above: It's clearly not "all"
> downstreams, as we don't carry anything similar to either of the two
> patches. But I'm pretty sure you meant "most" or "many" anyway.

Yes, indeed not all.  I don't carry those either on the FreeBSD
packages.

Thanks, Roger.