From: Alex Williamson <alex.williamson@redhat.com>
To: Feng Wu <feng.wu@intel.com>
Cc: pbonzini@redhat.com, joro@8bytes.org, mtosatti@redhat.com,
eric.auger@linaro.org, kvm@vger.kernel.org,
iommu@lists.linux-foundation.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH v9 12/18] vfio: Register/unregister irq_bypass_producer
Date: Tue, 26 Apr 2016 14:08:16 -0600 [thread overview]
Message-ID: <20160426140816.67b8b37c@t450s.home> (raw)
In-Reply-To: <1442586596-5920-13-git-send-email-feng.wu@intel.com>
On Fri, 18 Sep 2015 22:29:50 +0800
Feng Wu <feng.wu@intel.com> wrote:
@@ -360,6 +361,14 @@ static int vfio_msi_set_vector_signal(struct
vfio_pci_device *vdev,
> return ret;
> }
>
> + vdev->ctx[vector].producer.token = trigger;
> + vdev->ctx[vector].producer.irq = irq;
> + ret = irq_bypass_register_producer(&vdev->ctx[vector].producer);
> + if (unlikely(ret))
> + dev_info(&pdev->dev,
> + "irq bypass producer (token %p) registeration fails: %d\n",
> + vdev->ctx[vector].producer.token, ret);
> +
> vdev->ctx[vector].trigger = trigger;
>
> return 0;
Digging back into the IRQ producer/consumer thing, I'm not sure how we
should be handling a failure here, but it turns out that what we have
is pretty sub-optimal. Any sort of testing on AMD hits this dev_info
because kvm_arch_irq_bypass_add_producer() returns -EINVAL without
kvm_x86_ops->update_pi_irte which is only implemented for vmx. Clearly
we don't want to spew confusing error messages for a feature that does
not exist.
The easiest option is to simply make this error silent, but should
registering a producer/consumer really fail due to a mismatch on the
other end or should the __connect sequence fail silently, which both
ends would know about (if they care) due to the add/del handshake
between them? Perhaps for now we simply need a stable suitable fix to
silence the dev_info above, but longer term, registration shouldn't
fail for mismatches like this. Thoughts? Thanks,
Alex
WARNING: multiple messages have this Message-ID (diff)
From: Alex Williamson <alex.williamson-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
To: Feng Wu <feng.wu-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Cc: kvm-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
eric.auger-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org,
mtosatti-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org,
linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
iommu-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org,
pbonzini-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org
Subject: Re: [PATCH v9 12/18] vfio: Register/unregister irq_bypass_producer
Date: Tue, 26 Apr 2016 14:08:16 -0600 [thread overview]
Message-ID: <20160426140816.67b8b37c@t450s.home> (raw)
In-Reply-To: <1442586596-5920-13-git-send-email-feng.wu-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
On Fri, 18 Sep 2015 22:29:50 +0800
Feng Wu <feng.wu-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org> wrote:
@@ -360,6 +361,14 @@ static int vfio_msi_set_vector_signal(struct
vfio_pci_device *vdev,
> return ret;
> }
>
> + vdev->ctx[vector].producer.token = trigger;
> + vdev->ctx[vector].producer.irq = irq;
> + ret = irq_bypass_register_producer(&vdev->ctx[vector].producer);
> + if (unlikely(ret))
> + dev_info(&pdev->dev,
> + "irq bypass producer (token %p) registeration fails: %d\n",
> + vdev->ctx[vector].producer.token, ret);
> +
> vdev->ctx[vector].trigger = trigger;
>
> return 0;
Digging back into the IRQ producer/consumer thing, I'm not sure how we
should be handling a failure here, but it turns out that what we have
is pretty sub-optimal. Any sort of testing on AMD hits this dev_info
because kvm_arch_irq_bypass_add_producer() returns -EINVAL without
kvm_x86_ops->update_pi_irte which is only implemented for vmx. Clearly
we don't want to spew confusing error messages for a feature that does
not exist.
The easiest option is to simply make this error silent, but should
registering a producer/consumer really fail due to a mismatch on the
other end or should the __connect sequence fail silently, which both
ends would know about (if they care) due to the add/del handshake
between them? Perhaps for now we simply need a stable suitable fix to
silence the dev_info above, but longer term, registration shouldn't
fail for mismatches like this. Thoughts? Thanks,
Alex
next prev parent reply other threads:[~2016-04-26 20:08 UTC|newest]
Thread overview: 107+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-09-18 14:29 [PATCH v9 00/18] Add VT-d Posted-Interrupts support - including prerequisite series Feng Wu
2015-09-18 14:29 ` Feng Wu
2015-09-18 14:29 ` [PATCH v9 01/18] virt: IRQ bypass manager Feng Wu
2015-09-18 14:29 ` Feng Wu
2015-09-18 15:34 ` Wu, Feng
2015-09-18 15:34 ` Wu, Feng
2015-09-18 14:29 ` [PATCH v9 02/18] KVM: x86: select IRQ_BYPASS_MANAGER Feng Wu
2015-09-18 14:29 ` Feng Wu
2015-09-18 14:29 ` [PATCH v9 03/18] KVM: arm/arm64: " Feng Wu
2015-09-21 19:32 ` Eric Auger
2015-09-18 14:29 ` [PATCH v9 04/18] KVM: create kvm_irqfd.h Feng Wu
2015-09-18 14:29 ` Feng Wu
2015-09-18 15:35 ` Wu, Feng
2015-09-18 15:35 ` Wu, Feng
2015-09-18 14:29 ` [PATCH v9 05/18] KVM: introduce kvm_arch functions for IRQ bypass Feng Wu
2015-09-18 14:29 ` Feng Wu
2015-09-18 14:29 ` [PATCH v9 06/18] KVM: eventfd: add irq bypass consumer management Feng Wu
2015-09-18 14:29 ` Feng Wu
2015-09-18 14:29 ` [PATCH v9 07/18] KVM: Extend struct pi_desc for VT-d Posted-Interrupts Feng Wu
2015-09-18 14:29 ` Feng Wu
2015-09-18 14:29 ` [PATCH v9 08/18] KVM: Add some helper functions for Posted-Interrupts Feng Wu
2015-09-18 14:29 ` [PATCH v9 09/18] KVM: Define a new interface kvm_intr_is_single_vcpu() Feng Wu
2015-09-18 14:29 ` Feng Wu
2015-09-18 14:29 ` [PATCH v9 10/18] KVM: Make struct kvm_irq_routing_table accessible Feng Wu
2015-09-18 14:29 ` [PATCH v9 11/18] KVM: make kvm_set_msi_irq() public Feng Wu
2015-09-18 14:29 ` Feng Wu
2015-09-18 14:29 ` [PATCH v9 12/18] vfio: Register/unregister irq_bypass_producer Feng Wu
2015-09-18 14:29 ` Feng Wu
2015-09-18 17:19 ` Alex Williamson
2015-09-18 17:19 ` Alex Williamson
2015-09-21 8:56 ` Wu, Feng
2015-09-21 8:56 ` Wu, Feng
2015-09-21 9:32 ` Paolo Bonzini
2015-09-21 9:32 ` Paolo Bonzini
2015-09-21 11:35 ` Wu, Feng
2015-09-21 11:35 ` Wu, Feng
2015-09-21 12:06 ` Paolo Bonzini
2015-09-21 12:06 ` Paolo Bonzini
2015-09-21 12:08 ` Wu, Feng
2015-09-21 12:08 ` Wu, Feng
2015-09-21 12:53 ` Wu, Feng
2015-09-21 12:53 ` Wu, Feng
2015-09-21 13:02 ` Paolo Bonzini
2015-09-21 19:46 ` Eric Auger
2015-09-21 19:46 ` Eric Auger
2016-04-26 20:08 ` Alex Williamson [this message]
2016-04-26 20:08 ` Alex Williamson
2016-04-27 1:32 ` Wu, Feng
2016-04-27 1:32 ` Wu, Feng
2016-04-28 16:40 ` Alex Williamson
2016-04-28 16:40 ` Alex Williamson
2016-04-28 15:35 ` Eric Auger
2016-04-28 15:35 ` Eric Auger
2015-09-18 14:29 ` [PATCH v9 13/18] KVM: x86: Update IRTE for posted-interrupts Feng Wu
2015-09-18 14:29 ` Feng Wu
2015-09-18 14:29 ` [PATCH v9 14/18] KVM: Implement IRQ bypass consumer callbacks for x86 Feng Wu
2015-09-18 14:29 ` Feng Wu
2015-09-18 14:29 ` [PATCH v9 15/18] KVM: Add an arch specific hooks in 'struct kvm_kernel_irqfd' Feng Wu
2015-09-18 14:29 ` Feng Wu
2015-09-18 14:29 ` [PATCH v9 16/18] KVM: Update Posted-Interrupts Descriptor when vCPU is preempted Feng Wu
2015-09-18 14:29 ` Feng Wu
2015-09-18 14:29 ` [PATCH v9 17/18] KVM: Update Posted-Interrupts Descriptor when vCPU is blocked Feng Wu
2015-09-18 14:29 ` Feng Wu
2015-09-18 16:06 ` Paolo Bonzini
2015-09-18 16:06 ` Paolo Bonzini
2015-09-19 7:11 ` Wu, Feng
2015-09-19 7:11 ` Wu, Feng
2015-09-21 2:16 ` Wu, Feng
2015-09-21 5:32 ` Paolo Bonzini
2015-09-21 5:32 ` Paolo Bonzini
2015-09-21 5:45 ` Wu, Feng
2015-09-21 5:45 ` Wu, Feng
2015-10-14 23:41 ` David Matlack
2015-10-15 1:33 ` Wu, Feng
2015-10-15 1:33 ` Wu, Feng
2015-10-15 17:39 ` David Matlack
2015-10-15 18:13 ` Paolo Bonzini
2015-10-15 18:13 ` Paolo Bonzini
2015-10-16 1:45 ` Wu, Feng
2015-10-16 1:45 ` Wu, Feng
2015-09-18 14:29 ` [PATCH v9 18/18] iommu/vt-d: Add a command line parameter for VT-d posted-interrupts Feng Wu
2015-09-18 14:29 ` Feng Wu
2015-09-21 13:46 ` Joerg Roedel
2015-09-21 13:46 ` Joerg Roedel
2015-09-18 14:58 ` [PATCH v9 00/18] Add VT-d Posted-Interrupts support - including prerequisite series Paolo Bonzini
2015-09-18 14:58 ` Paolo Bonzini
2015-09-18 15:08 ` Wu, Feng
2015-09-18 15:08 ` Wu, Feng
2015-09-18 15:21 ` Paolo Bonzini
2015-09-18 15:21 ` Paolo Bonzini
2015-09-18 15:38 ` Wu, Feng
2015-09-18 17:57 ` Alex Williamson
2015-09-18 17:57 ` Alex Williamson
2015-09-25 1:49 ` Wu, Feng
2015-09-25 1:49 ` Wu, Feng
2015-09-25 11:14 ` Paolo Bonzini
2015-09-25 11:14 ` Paolo Bonzini
2015-09-28 10:14 ` Wu, Feng
2015-09-28 10:14 ` Wu, Feng
2015-09-28 10:18 ` Paolo Bonzini
2015-09-28 10:18 ` Paolo Bonzini
2015-09-28 10:22 ` Wu, Feng
2015-09-28 10:22 ` Wu, Feng
2015-09-22 9:01 [PATCH v9 12/18] vfio: Register/unregister irq_bypass_producer Wu, Feng
2015-09-22 9:01 ` Wu, Feng
2015-09-23 1:35 ` Eric Auger
2015-09-23 1:35 ` Eric Auger
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20160426140816.67b8b37c@t450s.home \
--to=alex.williamson@redhat.com \
--cc=eric.auger@linaro.org \
--cc=feng.wu@intel.com \
--cc=iommu@lists.linux-foundation.org \
--cc=joro@8bytes.org \
--cc=kvm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mtosatti@redhat.com \
--cc=pbonzini@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.