| Message ID | 20190919150314.054351477@linutronix.de (mailing list archive) |
|---|---|
| Headers | show |
| Series | entry: Provide generic implementation for host and guest entry/exit work | expand |
Hi Thomas, As a heads-up, I'm going to be away next week, and I likely won't have the chance to look at this in detail before October. On Thu, Sep 19, 2019 at 05:03:14PM +0200, Thomas Gleixner wrote: > When working on a way to move out the posix cpu timer expiry out of the > timer interrupt context, I noticed that KVM is not handling pending task > work before entering a guest. A quick hack was to add that to the x86 KVM > handling loop. The discussion ended with a request to make this a generic > infrastructure possible with also moving the per arch implementations of > the enter from and return to user space handling generic. > > https://lore.kernel.org/r/89E42BCC-47A8-458B-B06A-D6A20D20512C@amacapital.net > > You asked for it, so don't complain that you have to review it :) I never asked for this! ;) > The series implements the syscall enter/exit and the general exit to > userspace work handling along with the pre guest enter functionality. > > The series converts x86 and ARM64. x86 is fully tested including selftests > etc. ARM64 is only compile tested for now as my only ARM64 testbox is not > available right now. I've been working on converting the arm64 entry code to C for a while now [1], gradually upstreaming the bits I can. James has picked up some of that [2] as a prerequisite for some RAS error handling, and I think building the arm64 bits atop of that would be preferable. IIUC that should get posted as a series come -rc1. Since there's immense scope for subtle breakage, I'd prefer that we do the arm64-specific asm->C conversion before migrating arm64 to generic code. That way us arm64 folk can ensure the asm->C conversion retains the existing behaviour, and it'll be easier for everyone to compare the arm64 and generic C implementations. Thanks, Mark. [1] git://git.kernel.org/pub/scm/linux/kernel/git/mark/linux.git [2] git://linux-arm.org/linux-jm.git -b deasm_sync_only/v1 > > Thanks, > > tglx > > --- > /Makefile | 3 > arch/Kconfig | 3 > arch/arm64/Kconfig | 1 > arch/arm64/include/asm/kvm_host.h | 1 > arch/arm64/kernel/entry.S | 18 - > arch/arm64/kernel/ptrace.c | 65 ------ > arch/arm64/kernel/signal.c | 45 ---- > arch/arm64/kernel/syscall.c | 49 ---- > arch/x86/Kconfig | 1 > arch/x86/entry/common.c | 265 +------------------------- > arch/x86/entry/entry_32.S | 13 - > arch/x86/entry/entry_64.S | 12 - > arch/x86/entry/entry_64_compat.S | 21 -- > arch/x86/include/asm/signal.h | 1 > arch/x86/include/asm/thread_info.h | 9 > arch/x86/kernel/signal.c | 2 > arch/x86/kvm/x86.c | 17 - > b/arch/arm64/include/asm/entry-common.h | 76 +++++++ > b/arch/x86/include/asm/entry-common.h | 104 ++++++++++ > b/include/linux/entry-common.h | 324 ++++++++++++++++++++++++++++++++ > b/kernel/entry/common.c | 220 +++++++++++++++++++++ > kernel/Makefile | 1 > 22 files changed, 776 insertions(+), 475 deletions(-) > >
On Thu, Sep 19, 2019 at 8:09 AM Thomas Gleixner <tglx@linutronix.de> wrote: > > When working on a way to move out the posix cpu timer expiry out of the > timer interrupt context, I noticed that KVM is not handling pending task > work before entering a guest. A quick hack was to add that to the x86 KVM > handling loop. The discussion ended with a request to make this a generic > infrastructure possible with also moving the per arch implementations of > the enter from and return to user space handling generic. > > https://lore.kernel.org/r/89E42BCC-47A8-458B-B06A-D6A20D20512C@amacapital.net > > You asked for it, so don't complain that you have to review it :) > > The series implements the syscall enter/exit and the general exit to > userspace work handling along with the pre guest enter functionality. > > The series converts x86 and ARM64. x86 is fully tested including selftests > etc. ARM64 is only compile tested for now as my only ARM64 testbox is not > available right now. Other than the comments I sent so far, I like this series.
On Fri, 20 Sep 2019, Mark Rutland wrote: > I've been working on converting the arm64 entry code to C for a while > now [1], gradually upstreaming the bits I can. > > James has picked up some of that [2] as a prerequisite for some RAS > error handling, and I think building the arm64 bits atop of that would > be preferable. IIUC that should get posted as a series come -rc1. > > Since there's immense scope for subtle breakage, I'd prefer that we do > the arm64-specific asm->C conversion before migrating arm64 to generic > code. That way us arm64 folk can ensure the asm->C conversion retains > the existing behaviour, and it'll be easier for everyone to compare the > arm64 and generic C implementations. Right. It still would be nice to have some feedback on the general approach. That sais I'm happy to let you screw your entry code up yourself :) Thanks tglx
On 19.09.19 17:03, Thomas Gleixner wrote: > When working on a way to move out the posix cpu timer expiry out of the > timer interrupt context, I noticed that KVM is not handling pending task > work before entering a guest. A quick hack was to add that to the x86 KVM > handling loop. The discussion ended with a request to make this a generic > infrastructure possible with also moving the per arch implementations of > the enter from and return to user space handling generic. > > https://lore.kernel.org/r/89E42BCC-47A8-458B-B06A-D6A20D20512C@amacapital.net > > You asked for it, so don't complain that you have to review it :) > > The series implements the syscall enter/exit and the general exit to > userspace work handling along with the pre guest enter functionality. > > The series converts x86 and ARM64. x86 is fully tested including selftests > etc. ARM64 is only compile tested for now as my only ARM64 testbox is not > available right now. It seems that s390x would also need to look into TIF_NOTIFY_PENDING before entering a KVM guest. Given that the s390x entry path is still in assembler this might not be something to do quickly. Would it make sense to actually start with a minimal solution (e.g. one that provides notify_resume_pending like your original patch) as a fix. That would also be simple to backport. And then we can do the proper rework on top. Or do we consider anything that depends on TIF_NOTIFY_PENDING before entering a guest as not important enough for stable? After all the vcpu_run ioctl almost never returns to userspace and nothing was obviously broken. Another question: Are there callbacks due to TIF_NOTIFY_PENDING that should NOT happen as long as we stay in the vpcu loop?