mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
To: linux-kernel@vger.kernel.org
Cc: Jan Beulich <JBeulich@novell.com>,
	xen-devel@lists.xensource.com,
	Jeremy Fitzhardinge <jeremy@goop.org>,
	Konrad Rzeszutek Wilk <konrad@kernel.org>,
	Stefano Stabellini <stefano.stabellini@eu.citrix.com>,
	Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>,
	Weidong Han <weidong.han@intel.com>,
	Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
Subject: [PATCH 11/23] xen: fix shared irq device passthrough
Date: Tue, 12 Oct 2010 11:44:19 -0400	[thread overview]
Message-ID: <1286898271-32018-12-git-send-email-konrad.wilk@oracle.com> (raw)
In-Reply-To: <1286898271-32018-1-git-send-email-konrad.wilk@oracle.com>

In driver/xen/events.c, whether bind_pirq is shareable or not is
determined by desc->action is NULL or not. But in __setup_irq,
startup(irq) is invoked before desc->action is assigned with
new action. So desc->action in startup_irq is always NULL, and
bind_pirq is always not shareable. This results in pt_irq_create_bind
failure when passthrough a device which shares irq to other devices.

This patch doesn't use probing_irq to determine if pirq is shareable
or not, instead set shareable flag in irq_info according to trigger
mode in xen_allocate_pirq. Set level triggered interrupts shareable.
Thus use this flag to set bind_pirq flag accordingly.

[v2: arch/x86/xen/pci.c no more, so file skipped]

Signed-off-by: Weidong Han <weidong.han@intel.com>
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
---
 drivers/xen/events.c |   11 ++++++++---
 include/xen/events.h |    2 +-
 2 files changed, 9 insertions(+), 4 deletions(-)

diff --git a/drivers/xen/events.c b/drivers/xen/events.c
index 427f2d8..bf1dde6 100644
--- a/drivers/xen/events.c
+++ b/drivers/xen/events.c
@@ -97,6 +97,7 @@ struct irq_info
 	} u;
 };
 #define PIRQ_NEEDS_EOI	(1 << 0)
+#define PIRQ_SHAREABLE	(1 << 1)
 
 static struct irq_info *irq_info;
 
@@ -446,6 +447,7 @@ static unsigned int startup_pirq(unsigned int irq)
 	struct evtchn_bind_pirq bind_pirq;
 	struct irq_info *info = info_for_irq(irq);
 	int evtchn = evtchn_from_irq(irq);
+	int rc;
 
 	BUG_ON(info->type != IRQT_PIRQ);
 
@@ -454,8 +456,10 @@ static unsigned int startup_pirq(unsigned int irq)
 
 	bind_pirq.pirq = irq;
 	/* NB. We are happy to share unless we are probing. */
-	bind_pirq.flags = probing_irq(irq) ? 0 : BIND_PIRQ__WILL_SHARE;
-	if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_pirq, &bind_pirq) != 0) {
+	bind_pirq.flags = info->u.pirq.flags & PIRQ_SHAREABLE ?
+					BIND_PIRQ__WILL_SHARE : 0;
+	rc = HYPERVISOR_event_channel_op(EVTCHNOP_bind_pirq, &bind_pirq);
+	if (rc != 0) {
 		if (!probing_irq(irq))
 			printk(KERN_INFO "Failed to obtain physical IRQ %d\n",
 			       irq);
@@ -565,7 +569,7 @@ static int find_irq_by_gsi(unsigned gsi)
  * event channel until the irq actually started up.  Return an
  * existing irq if we've already got one for the gsi.
  */
-int xen_allocate_pirq(unsigned gsi, char *name)
+int xen_allocate_pirq(unsigned gsi, int shareable, char *name)
 {
 	int irq;
 	struct physdev_irq irq_op;
@@ -597,6 +601,7 @@ int xen_allocate_pirq(unsigned gsi, char *name)
 	}
 
 	irq_info[irq] = mk_pirq_info(0, gsi, irq_op.vector);
+	irq_info[irq].u.pirq.flags |= shareable ? PIRQ_SHAREABLE : 0;
 
 out:
 	spin_unlock(&irq_mapping_update_lock);
diff --git a/include/xen/events.h b/include/xen/events.h
index 2532f8b..d7a4ca7 100644
--- a/include/xen/events.h
+++ b/include/xen/events.h
@@ -70,7 +70,7 @@ void xen_hvm_evtchn_do_upcall(void);
 /* Allocate an irq for a physical interrupt, given a gsi.  "Legacy"
  * GSIs are identity mapped; others are dynamically allocated as
  * usual. */
-int xen_allocate_pirq(unsigned gsi, char *name);
+int xen_allocate_pirq(unsigned gsi, int shareable, char *name);
 
 /* Return vector allocated to pirq */
 int xen_vector_from_irq(unsigned pirq);
-- 
1.7.0.4


  parent reply	other threads:[~2010-10-12 15:47 UTC|newest]

Thread overview: 32+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-10-12 15:44 [PATCH v8] Xen PCI + Xen PCI frontend driver Konrad Rzeszutek Wilk
2010-10-12 15:44 ` [PATCH 01/23] xen: Don't disable the I/O space Konrad Rzeszutek Wilk
2010-10-12 15:44 ` [PATCH 02/23] xen: define BIOVEC_PHYS_MERGEABLE() Konrad Rzeszutek Wilk
2010-10-12 15:44 ` [PATCH 03/23] xen: implement pirq type event channels Konrad Rzeszutek Wilk
2010-10-12 15:44 ` [PATCH 04/23] x86/io_apic: add get_nr_irqs_gsi() Konrad Rzeszutek Wilk
2010-10-12 15:44 ` [PATCH 05/23] xen: identity map gsi->irqs Konrad Rzeszutek Wilk
2010-10-12 15:44 ` [PATCH 06/23] xen: dynamically allocate irq & event structures Konrad Rzeszutek Wilk
2010-10-12 15:44 ` [PATCH 07/23] xen: set pirq name to something useful Konrad Rzeszutek Wilk
2010-10-12 15:44 ` [PATCH 08/23] xen: statically initialize cpu_evtchn_mask_p Konrad Rzeszutek Wilk
2011-01-24 17:44   ` Paolo Bonzini
2011-01-25 14:02     ` [Xen-devel] " Ian Campbell
2010-10-12 15:44 ` [PATCH 09/23] xen: Find an unbound irq number in reverse order (high to low) Konrad Rzeszutek Wilk
2010-10-12 15:44 ` [PATCH 10/23] xen: Provide a variant of xen_poll_irq with timeout Konrad Rzeszutek Wilk
2010-10-12 15:44 ` Konrad Rzeszutek Wilk [this message]
2010-10-12 15:44 ` [PATCH 12/23] x86/PCI: Clean up pci_cache_line_size Konrad Rzeszutek Wilk
2010-10-12 15:44 ` [PATCH 13/23] x86/PCI: make sure _PAGE_IOMAP it set on pci mappings Konrad Rzeszutek Wilk
2010-10-12 15:54   ` Jesse Barnes
2010-10-12 15:44 ` [PATCH 14/23] x86/PCI: Export pci_walk_bus function Konrad Rzeszutek Wilk
2010-10-12 15:44 ` [PATCH 15/23] msi: Introduce default_[teardown|setup]_msi_irqs with fallback Konrad Rzeszutek Wilk
2010-10-12 15:44 ` [PATCH 16/23] x86: Introduce x86_msi_ops Konrad Rzeszutek Wilk
2010-10-12 15:44 ` [PATCH 17/23] xen/x86/PCI: Add support for the Xen PCI subsystem Konrad Rzeszutek Wilk
2010-10-12 15:44 ` [PATCH 18/23] xenbus: Xen paravirtualised PCI hotplug support Konrad Rzeszutek Wilk
2010-10-12 15:44 ` [PATCH 19/23] xenbus: prevent warnings on unhandled enumeration values Konrad Rzeszutek Wilk
2010-10-12 15:44 ` [PATCH 20/23] xen-pcifront: Xen PCI frontend driver Konrad Rzeszutek Wilk
2010-10-13  9:36   ` Jan Beulich
2010-10-13 13:53     ` Konrad Rzeszutek Wilk
2010-10-13 16:16       ` Konrad Rzeszutek Wilk
2010-10-14  7:15         ` [Xen-devel] " Jan Beulich
2010-10-14 17:35           ` Konrad Rzeszutek Wilk
2010-10-12 15:44 ` [PATCH 21/23] xen/pci: Request ACS when Xen-SWIOTLB is activated Konrad Rzeszutek Wilk
2010-10-12 15:44 ` [PATCH 22/23] MAINTAINERS: Add myself for Xen PCI and Xen SWIOTLB maintainer Konrad Rzeszutek Wilk
2010-10-12 15:44 ` [PATCH 23/23] swiotlb-xen: On x86-32 builts, select SWIOTLB instead of depending on it Konrad Rzeszutek Wilk

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=1286898271-32018-12-git-send-email-konrad.wilk@oracle.com \
    --to=konrad.wilk@oracle.com \
    --cc=JBeulich@novell.com \
    --cc=jeremy.fitzhardinge@citrix.com \
    --cc=jeremy@goop.org \
    --cc=konrad@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=stefano.stabellini@eu.citrix.com \
    --cc=weidong.han@intel.com \
    --cc=xen-devel@lists.xensource.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®