| Message ID | 20210601095838.GA838783@bogus |
|---|---|
| State | New |
| Headers | show |
| Series | [GIT,PULL] firmware: arm_ffa: Initial support for v5.14 | expand |
Hi Olof, On Tue, Jun 01, 2021 at 10:58:38AM +0100, Sudeep Holla wrote: > Hi ARM SoC Team, > > Please pull ! > > This is a new driver pull request. One of the arm64 patch is being > pulled from a stable arm64 branch for-next/ffa as the other patches > are dependent of the same. > > Background > ---------- > This has been on the list for almost a year now with changing requirements. > Initially Arm KVM wanted to use this via userspace interface in VMM to > communicate with VMs. But it was later dropped in favour of arch-agnostic > interface[1]. > > Also there was some discussion on the dt-bindings which was dropped > completely. Though we need to workaround the lack of full discoveribility > in v1.0 spec, it is now being fixed for the next version of the spec. > I see that you have pulled quite a lot of drivers and other SoC code including my scmi and juno ones that were sent more recently than this one. Just thought of checking if this is still in queue ? Sorry for the nag but this has been on a list almost a year. We missed v5.13 due to DT binding controveries(for good reasons) and we really don't want to miss v5.14
Hi, On Mon, Jun 14, 2021 at 06:08:56PM +0100, Sudeep Holla wrote: > Hi Olof, > > On Tue, Jun 01, 2021 at 10:58:38AM +0100, Sudeep Holla wrote: > > Hi ARM SoC Team, > > > > Please pull ! > > > > This is a new driver pull request. One of the arm64 patch is being > > pulled from a stable arm64 branch for-next/ffa as the other patches > > are dependent of the same. > > > > Background > > ---------- > > This has been on the list for almost a year now with changing requirements. > > Initially Arm KVM wanted to use this via userspace interface in VMM to > > communicate with VMs. But it was later dropped in favour of arch-agnostic > > interface[1]. > > > > Also there was some discussion on the dt-bindings which was dropped > > completely. Though we need to workaround the lack of full discoveribility > > in v1.0 spec, it is now being fixed for the next version of the spec. > > > > I see that you have pulled quite a lot of drivers and other SoC code > including my scmi and juno ones that were sent more recently than this > one. Just thought of checking if this is still in queue ? Sorry for the > nag but this has been on a list almost a year. We missed v5.13 due to > DT binding controveries(for good reasons) and we really don't want to > miss v5.14 I held off because I wanted to spend a few cycles on looking into it before blindly merging it. Are there any implemented users of this (on either side) today? We normally don't merge a framework in the kernel without having at least one user of it also available. -Olof
Hi Olof, On Tue, Jun 15, 2021 at 08:21:49AM -0700, Olof Johansson wrote: > Hi, > > On Mon, Jun 14, 2021 at 06:08:56PM +0100, Sudeep Holla wrote: > > Hi Olof, > > > > On Tue, Jun 01, 2021 at 10:58:38AM +0100, Sudeep Holla wrote: > > > Hi ARM SoC Team, > > > > > > Please pull ! > > > > > > This is a new driver pull request. One of the arm64 patch is being > > > pulled from a stable arm64 branch for-next/ffa as the other patches > > > are dependent of the same. > > > > > > Background > > > ---------- > > > This has been on the list for almost a year now with changing requirements. > > > Initially Arm KVM wanted to use this via userspace interface in VMM to > > > communicate with VMs. But it was later dropped in favour of arch-agnostic > > > interface[1]. > > > > > > Also there was some discussion on the dt-bindings which was dropped > > > completely. Though we need to workaround the lack of full discoveribility > > > in v1.0 spec, it is now being fixed for the next version of the spec. > > > > > > > I see that you have pulled quite a lot of drivers and other SoC code > > including my scmi and juno ones that were sent more recently than this > > one. Just thought of checking if this is still in queue ? Sorry for the > > nag but this has been on a list almost a year. We missed v5.13 due to > > DT binding controveries(for good reasons) and we really don't want to > > miss v5.14 > > I held off because I wanted to spend a few cycles on looking into it before > blindly merging it. > Sure, just wanted to make sure it was in your list. > Are there any implemented users of this (on either side) today? We normally > don't merge a framework in the kernel without having at least one user of it > also available. > Fair enough. Yes, OPTEE patches are on the list[1]. Just to avoid complexity in merging Jens(OPTEE maintainer) held it off for next cycle allowing more time. But we have been testing all the versions of this driver posted on the list with OPTEE. There are other users too but they need userspace interface which is still work in progress. -- Regards, Sudeep [1] https://lore.kernel.org/r/20210527081404.1433177-1-jens.wiklander@linaro.org/
On Tue, Jun 15, 2021 at 04:34:30PM +0100, Sudeep Holla wrote: > Hi Olof, > > On Tue, Jun 15, 2021 at 08:21:49AM -0700, Olof Johansson wrote: > > Hi, > > > > On Mon, Jun 14, 2021 at 06:08:56PM +0100, Sudeep Holla wrote: > > > Hi Olof, > > > > > > On Tue, Jun 01, 2021 at 10:58:38AM +0100, Sudeep Holla wrote: > > > > Hi ARM SoC Team, > > > > > > > > Please pull ! > > > > > > > > This is a new driver pull request. One of the arm64 patch is being > > > > pulled from a stable arm64 branch for-next/ffa as the other patches > > > > are dependent of the same. > > > > > > > > Background > > > > ---------- > > > > This has been on the list for almost a year now with changing requirements. > > > > Initially Arm KVM wanted to use this via userspace interface in VMM to > > > > communicate with VMs. But it was later dropped in favour of arch-agnostic > > > > interface[1]. > > > > > > > > Also there was some discussion on the dt-bindings which was dropped > > > > completely. Though we need to workaround the lack of full discoveribility > > > > in v1.0 spec, it is now being fixed for the next version of the spec. > > > > > > > > > > I see that you have pulled quite a lot of drivers and other SoC code > > > including my scmi and juno ones that were sent more recently than this > > > one. Just thought of checking if this is still in queue ? Sorry for the > > > nag but this has been on a list almost a year. We missed v5.13 due to > > > DT binding controveries(for good reasons) and we really don't want to > > > miss v5.14 > > > > I held off because I wanted to spend a few cycles on looking into it before > > blindly merging it. > > > > Sure, just wanted to make sure it was in your list. Yeah, thanks for the ping. Whenever we miss something we almost always say: "But you didn't ping us?!". You did, and it'd be awfully bad signaling if we somehow got upset by it if we want people to do it more often. > > Are there any implemented users of this (on either side) today? We normally > > don't merge a framework in the kernel without having at least one user of it > > also available. > > > > Fair enough. > > Yes, OPTEE patches are on the list[1]. Just to avoid complexity in merging > Jens(OPTEE maintainer) held it off for next cycle allowing more time. But we > have been testing all the versions of this driver posted on the list with > OPTEE. There are other users too but they need userspace interface which > is still work in progress. Ok, thanks for that! Ideally, having that in the tag, i.e. also in cover letter for patch set would have avoided this round trip next time. But thanks for that detail. I'll queue this up shortly. -Olof
On Tue, Jun 15, 2021 at 09:47:24AM -0700, Olof Johansson wrote: > On Tue, Jun 15, 2021 at 04:34:30PM +0100, Sudeep Holla wrote: > > Hi Olof, > > > > On Tue, Jun 15, 2021 at 08:21:49AM -0700, Olof Johansson wrote: > > > Hi, > > > > > > On Mon, Jun 14, 2021 at 06:08:56PM +0100, Sudeep Holla wrote: > > > > Hi Olof, > > > > > > > > On Tue, Jun 01, 2021 at 10:58:38AM +0100, Sudeep Holla wrote: > > > > > Hi ARM SoC Team, > > > > > > > > > > Please pull ! > > > > > > > > > > This is a new driver pull request. One of the arm64 patch is being > > > > > pulled from a stable arm64 branch for-next/ffa as the other patches > > > > > are dependent of the same. > > > > > > > > > > Background > > > > > ---------- > > > > > This has been on the list for almost a year now with changing requirements. > > > > > Initially Arm KVM wanted to use this via userspace interface in VMM to > > > > > communicate with VMs. But it was later dropped in favour of arch-agnostic > > > > > interface[1]. > > > > > > > > > > Also there was some discussion on the dt-bindings which was dropped > > > > > completely. Though we need to workaround the lack of full discoveribility > > > > > in v1.0 spec, it is now being fixed for the next version of the spec. > > > > > > > > > > > > > I see that you have pulled quite a lot of drivers and other SoC code > > > > including my scmi and juno ones that were sent more recently than this > > > > one. Just thought of checking if this is still in queue ? Sorry for the > > > > nag but this has been on a list almost a year. We missed v5.13 due to > > > > DT binding controveries(for good reasons) and we really don't want to > > > > miss v5.14 > > > > > > I held off because I wanted to spend a few cycles on looking into it before > > > blindly merging it. > > > > > > > Sure, just wanted to make sure it was in your list. > > Yeah, thanks for the ping. Whenever we miss something we almost always say: > "But you didn't ping us?!". You did, and it'd be awfully bad signaling if we > somehow got upset by it if we want people to do it more often. > Understood, I generally ping observing the pattern of pull requests being merged. Initially I thought you might be in LIFO mode but then observed you had even merged other PR around that date. So thought of checking and I do understand I don't want to annoy repeating that too much. It rarely happens but once the PR got marked read/done accidentally. Also the main reason for nagging on this particular one is it missed v5.13 and would help if it lands in upstream for "earlier" adoption(mostly dependent on android though). > > > Are there any implemented users of this (on either side) today? We normally > > > don't merge a framework in the kernel without having at least one user of it > > > also available. > > > > > > > Fair enough. > > > > Yes, OPTEE patches are on the list[1]. Just to avoid complexity in merging > > Jens(OPTEE maintainer) held it off for next cycle allowing more time. But we > > have been testing all the versions of this driver posted on the list with > > OPTEE. There are other users too but they need userspace interface which > > is still work in progress. > > Ok, thanks for that! Ideally, having that in the tag, i.e. also in cover letter > for patch set would have avoided this round trip next time. But thanks for that > detail. > Agreed, it seem to have missed throughout. I just used to mention that it is tested with OPTEE in each revision but never referred back to the series as it was always catch with OPTEE, every time I revised, OPTEE needed to rebase and I couldn't share something ready. My bad, may be we never synchronised our work that well. Point taken for any such future dependencies/references. > I'll queue this up shortly. > Thanks for that, much appreciated!
Hi ARM SoC Team, Please pull ! This is a new driver pull request. One of the arm64 patch is being pulled from a stable arm64 branch for-next/ffa as the other patches are dependent of the same. Background ---------- This has been on the list for almost a year now with changing requirements. Initially Arm KVM wanted to use this via userspace interface in VMM to communicate with VMs. But it was later dropped in favour of arch-agnostic interface[1]. Also there was some discussion on the dt-bindings which was dropped completely. Though we need to workaround the lack of full discoveribility in v1.0 spec, it is now being fixed for the next version of the spec. Regards, Sudeep [1] https://mirrors.edge.kernel.org/pub/linux/kernel/people/will/slides/kvmforum-2020-edited.pdf -->8 The following changes since commit 6efb943b8616ec53a5e444193dccf1af9ad627b5: Linux 5.13-rc1 (2021-05-09 14:17:44 -0700) are available in the Git repository at: git://git.kernel.org/pub/scm/linux/kernel/git/sudeep.holla/linux.git tags/arm-ffa-5.14 for you to fetch changes up to cc2195fe536c28e192df5d07e6dd277af36814b4: firmware: arm_ffa: Add support for MEM_* interfaces (2021-05-26 22:38:43 +0100) ---------------------------------------------------------------- Arm Firmware Framework for ARMv8-A(FFA) interface driver The Arm FFA specification describes a software architecture to leverages the virtualization extension to isolate software images provided by an ecosystem of vendors from each other and describes interfaces that standardize communication between the various software images including communication between images in the Secure world and Normal world. Any Hypervisor could use the FFA interfaces to enable communication between VMs it manages. The Hypervisor a.k.a Partition managers in FFA terminology can assign system resources(Memory regions, Devices, CPU cycles) to the partitions and manage isolation amongst them. This is the initial and minimal support for the FFA interface to enable communication between secure partitions and the normal world OS. ---------------------------------------------------------------- Sudeep Holla (6): arm64: smccc: Add support for SMCCCv1.2 extended input/output registers firmware: arm_ffa: Add initial FFA bus support for device enumeration firmware: arm_ffa: Add initial Arm FFA driver support firmware: arm_ffa: Add support for SMCCC as transport to FFA driver firmware: arm_ffa: Setup in-kernel users of FFA partitions firmware: arm_ffa: Add support for MEM_* interfaces MAINTAINERS | 7 + arch/arm64/kernel/asm-offsets.c | 9 + arch/arm64/kernel/smccc-call.S | 57 +++ drivers/firmware/Kconfig | 1 + drivers/firmware/Makefile | 1 + drivers/firmware/arm_ffa/Kconfig | 21 ++ drivers/firmware/arm_ffa/Makefile | 6 + drivers/firmware/arm_ffa/bus.c | 210 +++++++++++ drivers/firmware/arm_ffa/common.h | 31 ++ drivers/firmware/arm_ffa/driver.c | 731 ++++++++++++++++++++++++++++++++++++++ drivers/firmware/arm_ffa/smccc.c | 39 ++ include/linux/arm-smccc.h | 55 +++ include/linux/arm_ffa.h | 267 ++++++++++++++ 13 files changed, 1435 insertions(+) create mode 100644 drivers/firmware/arm_ffa/Kconfig create mode 100644 drivers/firmware/arm_ffa/Makefile create mode 100644 drivers/firmware/arm_ffa/bus.c create mode 100644 drivers/firmware/arm_ffa/common.h create mode 100644 drivers/firmware/arm_ffa/driver.c create mode 100644 drivers/firmware/arm_ffa/smccc.c create mode 100644 include/linux/arm_ffa.h