[SRU,B,D,0/9] zKVM: Add hardware CPU Model - kernel part (LP: 1836153)
mbox series

Message ID 1563260079-14121-1-git-send-email-frank.heimes@canonical.com
Headers show
Series
  • zKVM: Add hardware CPU Model - kernel part (LP: 1836153)
Related show

Message

Frank Heimes July 16, 2019, 6:54 a.m. UTC
Buglink: https://bugs.launchpad.net/bugs/1836153

SRU Justification:

[Impact]

* New hardware is not supported by qemu and won't run run, due to missing hardware CPU model

[Fix]

* 11ba5961a2156a4f210627ed8421387e2531b100 11ba596 "KVM: s390: add debug logging for cpu model subfunctions"

* 346fa2f891c71a9b98014f8f62c15f4c7dd95ec1 346fa2f "KVM: s390: implement subfunction processor calls"

* 7832e91cd33f21f3cf82b003478c292915a1ec14 7832e91 "KVM: s390: add vector enhancements facility 2 to cpumodel"

* d5cb6ab1e3d4d7e0648a167f6290e89f6e86964e d5cb6ab "KVM: s390: add vector BCD enhancements facility to cpumodel"

* 13209ad0395c4de7fa48108b1dac72e341d5c089 13209ad "KVM: s390: add MSA9 to cpumodel"

* d668139718a9e2260702777bd8d86d71c30b6539 d668139 "KVM: s390: provide query function for instructions returning 32 byte"

* 173aec2d5a9fa5f40e462661a8283fcafe04764f 173aec2 "KVM: s390: add enhanced sort facilty to cpu model"

* 4f45b90e1c03466202fca7f62eaf32243f220830 4f45b90 "KVM: s390: add deflate conversion facilty to cpu model"

* 8ec2fa52eac53bff7ef1cedbc4ad8af650ec937c 8ec2fa5 "KVM: s390: enable MSA9 keywrapping functions depending on cpu model"

[Test Case]

* need to be tested by IBM on pre-rel. hardware or simulator

[Regression Potential] 

* The regression potential can be considered as low since these changes are limited to arch/s390

* changes are in support for new and upcoming hardware and shouldn't affect existing s390x systems

[Other Info]

* the first 2 commits are the key ones, the other 7 are needed to make them applly cleanly

* these patches are already all included in upstream kernel 5.2, hence they are in eoan's kernel 5.2

* qemu package SRU in LP 1836154 complements this SRU - for testing both need to be in place

* commit IDs applied cleanly with 'cherry-picks --strategy=recursive -X theirs' - checked that no patches on top were pulled in

Christian Borntraeger (9):
  From: Christian Borntraeger <borntraeger@de.ibm.com>
  From: Christian Borntraeger <borntraeger@de.ibm.com>
  From: Christian Borntraeger <borntraeger@de.ibm.com>
  From: Christian Borntraeger <borntraeger@de.ibm.com>
  From: Christian Borntraeger <borntraeger@de.ibm.com>
  From: Christian Borntraeger <borntraeger@de.ibm.com>
  From: Christian Borntraeger <borntraeger@de.ibm.com>
  From: Christian Borntraeger <borntraeger@de.ibm.com>
  From: Christian Borntraeger <borntraeger@de.ibm.com>

 Documentation/virtual/kvm/devices/vm.txt |   3 +-
 arch/s390/include/asm/cpacf.h            |   1 +
 arch/s390/include/asm/kvm_host.h         |   2 +
 arch/s390/include/uapi/asm/kvm.h         |   5 +-
 arch/s390/kvm/kvm-s390.c                 | 271 +++++++++++++++++++++++++++++--
 arch/s390/kvm/vsie.c                     |   5 +-
 arch/s390/tools/gen_facilities.c         |   3 +
 tools/arch/s390/include/uapi/asm/kvm.h   |   3 +-
 8 files changed, 275 insertions(+), 18 deletions(-)

Comments

Stefan Bader July 18, 2019, 7:34 a.m. UTC | #1
On 16.07.19 08:54, frank.heimes@canonical.com wrote:
> Buglink: https://bugs.launchpad.net/bugs/1836153
> 
> SRU Justification:
> 
> [Impact]
> 
> * New hardware is not supported by qemu and won't run run, due to missing hardware CPU model
> 
> [Fix]
> 
> * 11ba5961a2156a4f210627ed8421387e2531b100 11ba596 "KVM: s390: add debug logging for cpu model subfunctions"
> 
> * 346fa2f891c71a9b98014f8f62c15f4c7dd95ec1 346fa2f "KVM: s390: implement subfunction processor calls"
> 
> * 7832e91cd33f21f3cf82b003478c292915a1ec14 7832e91 "KVM: s390: add vector enhancements facility 2 to cpumodel"
> 
> * d5cb6ab1e3d4d7e0648a167f6290e89f6e86964e d5cb6ab "KVM: s390: add vector BCD enhancements facility to cpumodel"
> 
> * 13209ad0395c4de7fa48108b1dac72e341d5c089 13209ad "KVM: s390: add MSA9 to cpumodel"
> 
> * d668139718a9e2260702777bd8d86d71c30b6539 d668139 "KVM: s390: provide query function for instructions returning 32 byte"
> 
> * 173aec2d5a9fa5f40e462661a8283fcafe04764f 173aec2 "KVM: s390: add enhanced sort facilty to cpu model"
> 
> * 4f45b90e1c03466202fca7f62eaf32243f220830 4f45b90 "KVM: s390: add deflate conversion facilty to cpu model"
> 
> * 8ec2fa52eac53bff7ef1cedbc4ad8af650ec937c 8ec2fa5 "KVM: s390: enable MSA9 keywrapping functions depending on cpu model"
> 
> [Test Case]
> 
> * need to be tested by IBM on pre-rel. hardware or simulator
> 
> [Regression Potential] 
> 
> * The regression potential can be considered as low since these changes are limited to arch/s390
> 
> * changes are in support for new and upcoming hardware and shouldn't affect existing s390x systems
> 
> [Other Info]
> 
> * the first 2 commits are the key ones, the other 7 are needed to make them applly cleanly
> 
> * these patches are already all included in upstream kernel 5.2, hence they are in eoan's kernel 5.2
> 
> * qemu package SRU in LP 1836154 complements this SRU - for testing both need to be in place
> 
> * commit IDs applied cleanly with 'cherry-picks --strategy=recursive -X theirs' - checked that no patches on top were pulled in
> 
> Christian Borntraeger (9):
>   From: Christian Borntraeger <borntraeger@de.ibm.com>
>   From: Christian Borntraeger <borntraeger@de.ibm.com>
>   From: Christian Borntraeger <borntraeger@de.ibm.com>
>   From: Christian Borntraeger <borntraeger@de.ibm.com>
>   From: Christian Borntraeger <borntraeger@de.ibm.com>
>   From: Christian Borntraeger <borntraeger@de.ibm.com>
>   From: Christian Borntraeger <borntraeger@de.ibm.com>
>   From: Christian Borntraeger <borntraeger@de.ibm.com>
>   From: Christian Borntraeger <borntraeger@de.ibm.com>
> 
>  Documentation/virtual/kvm/devices/vm.txt |   3 +-
>  arch/s390/include/asm/cpacf.h            |   1 +
>  arch/s390/include/asm/kvm_host.h         |   2 +
>  arch/s390/include/uapi/asm/kvm.h         |   5 +-
>  arch/s390/kvm/kvm-s390.c                 | 271 +++++++++++++++++++++++++++++--
>  arch/s390/kvm/vsie.c                     |   5 +-
>  arch/s390/tools/gen_facilities.c         |   3 +
>  tools/arch/s390/include/uapi/asm/kvm.h   |   3 +-
>  8 files changed, 275 insertions(+), 18 deletions(-)
> 
Not quite sure how you produced this but ^ it seems to have decided to use the
from line instead of the subject (there is a git format-patch --cover). But also
(the limit is floating) you might want to consider using pull requests when
there are more than 5 patches in a set.

Overall it seems a lot of the change is putting sections in and taking them out
again and then put them back in later. So real change would be less. And mostly
it is filling in things where only stubs were before. So the regression chance
is lower. And since it is limited to one arch:

Acked-by: Stefan Bader <stefan.bader@canonical.com>
Kleber Souza July 23, 2019, 10:36 a.m. UTC | #2
On 7/16/19 8:54 AM, frank.heimes@canonical.com wrote:
> Buglink: https://bugs.launchpad.net/bugs/1836153
> 
> SRU Justification:
> 
> [Impact]
> 
> * New hardware is not supported by qemu and won't run run, due to missing hardware CPU model
> 
> [Fix]
> 
> * 11ba5961a2156a4f210627ed8421387e2531b100 11ba596 "KVM: s390: add debug logging for cpu model subfunctions"
> 
> * 346fa2f891c71a9b98014f8f62c15f4c7dd95ec1 346fa2f "KVM: s390: implement subfunction processor calls"
> 
> * 7832e91cd33f21f3cf82b003478c292915a1ec14 7832e91 "KVM: s390: add vector enhancements facility 2 to cpumodel"
> 
> * d5cb6ab1e3d4d7e0648a167f6290e89f6e86964e d5cb6ab "KVM: s390: add vector BCD enhancements facility to cpumodel"
> 
> * 13209ad0395c4de7fa48108b1dac72e341d5c089 13209ad "KVM: s390: add MSA9 to cpumodel"
> 
> * d668139718a9e2260702777bd8d86d71c30b6539 d668139 "KVM: s390: provide query function for instructions returning 32 byte"
> 
> * 173aec2d5a9fa5f40e462661a8283fcafe04764f 173aec2 "KVM: s390: add enhanced sort facilty to cpu model"
> 
> * 4f45b90e1c03466202fca7f62eaf32243f220830 4f45b90 "KVM: s390: add deflate conversion facilty to cpu model"
> 
> * 8ec2fa52eac53bff7ef1cedbc4ad8af650ec937c 8ec2fa5 "KVM: s390: enable MSA9 keywrapping functions depending on cpu model"
> 
> [Test Case]
> 
> * need to be tested by IBM on pre-rel. hardware or simulator
> 
> [Regression Potential] 
> 
> * The regression potential can be considered as low since these changes are limited to arch/s390
> 
> * changes are in support for new and upcoming hardware and shouldn't affect existing s390x systems
> 
> [Other Info]
> 
> * the first 2 commits are the key ones, the other 7 are needed to make them applly cleanly
> 
> * these patches are already all included in upstream kernel 5.2, hence they are in eoan's kernel 5.2
> 
> * qemu package SRU in LP 1836154 complements this SRU - for testing both need to be in place
> 
> * commit IDs applied cleanly with 'cherry-picks --strategy=recursive -X theirs' - checked that no patches on top were pulled in
> 
> Christian Borntraeger (9):
>   From: Christian Borntraeger <borntraeger@de.ibm.com>
>   From: Christian Borntraeger <borntraeger@de.ibm.com>
>   From: Christian Borntraeger <borntraeger@de.ibm.com>
>   From: Christian Borntraeger <borntraeger@de.ibm.com>
>   From: Christian Borntraeger <borntraeger@de.ibm.com>
>   From: Christian Borntraeger <borntraeger@de.ibm.com>
>   From: Christian Borntraeger <borntraeger@de.ibm.com>
>   From: Christian Borntraeger <borntraeger@de.ibm.com>
>   From: Christian Borntraeger <borntraeger@de.ibm.com>
> 
>  Documentation/virtual/kvm/devices/vm.txt |   3 +-
>  arch/s390/include/asm/cpacf.h            |   1 +
>  arch/s390/include/asm/kvm_host.h         |   2 +
>  arch/s390/include/uapi/asm/kvm.h         |   5 +-
>  arch/s390/kvm/kvm-s390.c                 | 271 +++++++++++++++++++++++++++++--
>  arch/s390/kvm/vsie.c                     |   5 +-
>  arch/s390/tools/gen_facilities.c         |   3 +
>  tools/arch/s390/include/uapi/asm/kvm.h   |   3 +-
>  8 files changed, 275 insertions(+), 18 deletions(-)
> 

As Stefan mentioned, a lot of code moving around but probably easier to apply those
commit to have clean cherry picks later.

Changes specific to architecture and should affect only new platforms, so regression
potential looks low.

Acked-by: Kleber Sacilotto de Souza <kleber.souza@canonical.com>

Thank you,
Kleber
Kleber Souza July 23, 2019, 12:10 p.m. UTC | #3
On 7/16/19 8:54 AM, frank.heimes@canonical.com wrote:
> Buglink: https://bugs.launchpad.net/bugs/1836153
> 
> SRU Justification:
> 
> [Impact]
> 
> * New hardware is not supported by qemu and won't run run, due to missing hardware CPU model
> 
> [Fix]
> 
> * 11ba5961a2156a4f210627ed8421387e2531b100 11ba596 "KVM: s390: add debug logging for cpu model subfunctions"
> 
> * 346fa2f891c71a9b98014f8f62c15f4c7dd95ec1 346fa2f "KVM: s390: implement subfunction processor calls"
> 
> * 7832e91cd33f21f3cf82b003478c292915a1ec14 7832e91 "KVM: s390: add vector enhancements facility 2 to cpumodel"
> 
> * d5cb6ab1e3d4d7e0648a167f6290e89f6e86964e d5cb6ab "KVM: s390: add vector BCD enhancements facility to cpumodel"
> 
> * 13209ad0395c4de7fa48108b1dac72e341d5c089 13209ad "KVM: s390: add MSA9 to cpumodel"
> 
> * d668139718a9e2260702777bd8d86d71c30b6539 d668139 "KVM: s390: provide query function for instructions returning 32 byte"
> 
> * 173aec2d5a9fa5f40e462661a8283fcafe04764f 173aec2 "KVM: s390: add enhanced sort facilty to cpu model"
> 
> * 4f45b90e1c03466202fca7f62eaf32243f220830 4f45b90 "KVM: s390: add deflate conversion facilty to cpu model"
> 
> * 8ec2fa52eac53bff7ef1cedbc4ad8af650ec937c 8ec2fa5 "KVM: s390: enable MSA9 keywrapping functions depending on cpu model"
> 
> [Test Case]
> 
> * need to be tested by IBM on pre-rel. hardware or simulator
> 
> [Regression Potential] 
> 
> * The regression potential can be considered as low since these changes are limited to arch/s390
> 
> * changes are in support for new and upcoming hardware and shouldn't affect existing s390x systems
> 
> [Other Info]
> 
> * the first 2 commits are the key ones, the other 7 are needed to make them applly cleanly
> 
> * these patches are already all included in upstream kernel 5.2, hence they are in eoan's kernel 5.2
> 
> * qemu package SRU in LP 1836154 complements this SRU - for testing both need to be in place
> 
> * commit IDs applied cleanly with 'cherry-picks --strategy=recursive -X theirs' - checked that no patches on top were pulled in
> 
> Christian Borntraeger (9):
>   From: Christian Borntraeger <borntraeger@de.ibm.com>
>   From: Christian Borntraeger <borntraeger@de.ibm.com>
>   From: Christian Borntraeger <borntraeger@de.ibm.com>
>   From: Christian Borntraeger <borntraeger@de.ibm.com>
>   From: Christian Borntraeger <borntraeger@de.ibm.com>
>   From: Christian Borntraeger <borntraeger@de.ibm.com>
>   From: Christian Borntraeger <borntraeger@de.ibm.com>
>   From: Christian Borntraeger <borntraeger@de.ibm.com>
>   From: Christian Borntraeger <borntraeger@de.ibm.com>
> 
>  Documentation/virtual/kvm/devices/vm.txt |   3 +-
>  arch/s390/include/asm/cpacf.h            |   1 +
>  arch/s390/include/asm/kvm_host.h         |   2 +
>  arch/s390/include/uapi/asm/kvm.h         |   5 +-
>  arch/s390/kvm/kvm-s390.c                 | 271 +++++++++++++++++++++++++++++--
>  arch/s390/kvm/vsie.c                     |   5 +-
>  arch/s390/tools/gen_facilities.c         |   3 +
>  tools/arch/s390/include/uapi/asm/kvm.h   |   3 +-
>  8 files changed, 275 insertions(+), 18 deletions(-)
> 

Applied to bionic and disco master-next branches.

Thanks,
Kleber