From: Marc Zyngier <marc.zyngier@arm.com>
To: linux-kernel@vger.kernel.org,
linux-arm-kernel@lists.infradead.org,
kvmarm@lists.cs.columbia.edu
Cc: Christoffer Dall <christoffer.dall@linaro.org>,
Thomas Gleixner <tglx@linutronix.de>,
Jason Cooper <jason@lakedaemon.net>,
Eric Auger <eric.auger@redhat.com>,
Shanker Donthineni <shankerd@codeaurora.org>,
Mark Rutland <mark.rutland@arm.com>
Subject: [PATCH v2 18/52] irqchip/gic-v3-its: Add VLPI configuration handling
Date: Wed, 28 Jun 2017 16:03:37 +0100 [thread overview]
Message-ID: <20170628150411.15846-19-marc.zyngier@arm.com> (raw)
In-Reply-To: <20170628150411.15846-1-marc.zyngier@arm.com>
When a VLPI is reconfigured (enabled, disabled, change in priority),
the full configuration byte must be written, and the caches invalidated.
Also, when using the irq_mask/irq_unmask methods, it is necessary
to disable the doorbell for that particular interrupt (by mapping it
to 1023) on top of clearing the Enable bit.
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
drivers/irqchip/irq-gic-v3-its.c | 73 +++++++++++++++++++++++++++++++++++++---
1 file changed, 69 insertions(+), 4 deletions(-)
diff --git a/drivers/irqchip/irq-gic-v3-its.c b/drivers/irqchip/irq-gic-v3-its.c
index 527e4e95e10d..2e5e0499a0e6 100644
--- a/drivers/irqchip/irq-gic-v3-its.c
+++ b/drivers/irqchip/irq-gic-v3-its.c
@@ -808,18 +808,26 @@ static inline u32 its_get_event_id(struct irq_data *d)
return d->hwirq - its_dev->event_map.lpi_base;
}
-static void lpi_update_config(struct irq_data *d, u8 clr, u8 set)
+static void lpi_write_config(struct irq_data *d, u8 clr, u8 set)
{
struct its_device *its_dev = irq_data_get_irq_chip_data(d);
- irq_hw_number_t hwirq = d->hwirq;
+ irq_hw_number_t hwirq;
struct page *prop_page;
u8 *cfg;
- prop_page = gic_rdists->prop_page;
+ if (irqd_is_forwarded_to_vcpu(d)) {
+ u32 event = its_get_event_id(d);
+
+ prop_page = its_dev->event_map.vm->vprop_page;
+ hwirq = its_dev->event_map.vlpi_maps[event].vintid;
+ } else {
+ prop_page = gic_rdists->prop_page;
+ hwirq = d->hwirq;
+ }
cfg = page_address(prop_page) + hwirq - 8192;
*cfg &= ~clr;
- *cfg |= set;
+ *cfg |= set | LPI_PROP_GROUP1;
/*
* Make the above write visible to the redistributors.
@@ -830,16 +838,52 @@ static void lpi_update_config(struct irq_data *d, u8 clr, u8 set)
gic_flush_dcache_to_poc(cfg, sizeof(*cfg));
else
dsb(ishst);
+}
+
+static void lpi_update_config(struct irq_data *d, u8 clr, u8 set)
+{
+ struct its_device *its_dev = irq_data_get_irq_chip_data(d);
+
+ lpi_write_config(d, clr, set);
its_send_inv(its_dev, its_get_event_id(d));
}
+static void its_vlpi_set_doorbell(struct irq_data *d, bool enable)
+{
+ struct its_device *its_dev = irq_data_get_irq_chip_data(d);
+ u32 event = its_get_event_id(d);
+
+ if (its_dev->event_map.vlpi_maps[event].db_enabled == enable)
+ return;
+
+ its_dev->event_map.vlpi_maps[event].db_enabled = enable;
+
+ /*
+ * More fun with the architecture:
+ *
+ * Ideally, we'd issue a VMAPTI to set the doorbell to its LPI
+ * value or to 1023, depending on the enable bit. But that
+ * would be issueing a mapping for an /existing/ DevID+EventID
+ * pair, which is UNPREDICTABLE. Instead, let's issue a VMOVI
+ * to the /same/ vPE, using this opportunity to adjust the
+ * doorbell. Mouahahahaha. We loves it, Precious.
+ */
+ its_send_vmovi(its_dev, event);
+}
+
static void its_mask_irq(struct irq_data *d)
{
+ if (irqd_is_forwarded_to_vcpu(d))
+ its_vlpi_set_doorbell(d, false);
+
lpi_update_config(d, LPI_PROP_ENABLED, 0);
}
static void its_unmask_irq(struct irq_data *d)
{
+ if (irqd_is_forwarded_to_vcpu(d))
+ its_vlpi_set_doorbell(d, true);
+
lpi_update_config(d, 0, LPI_PROP_ENABLED);
}
@@ -852,6 +896,10 @@ static int its_set_affinity(struct irq_data *d, const struct cpumask *mask_val,
struct its_collection *target_col;
u32 id = its_get_event_id(d);
+ /* A forwarded interrupt should use irq_set_vcpu_affinity */
+ if (irqd_is_forwarded_to_vcpu(d))
+ return -EINVAL;
+
/* lpi cannot be routed to a redistributor that is on a foreign node */
if (its_dev->its->flags & ITS_FLAGS_WORKAROUND_CAVIUM_23144) {
if (its_dev->its->numa_node >= 0) {
@@ -1017,6 +1065,22 @@ static int its_vlpi_unmap(struct irq_data *d)
return ret;
}
+static int its_vlpi_prop_update(struct irq_data *d, struct its_cmd_info *info)
+{
+ struct its_device *its_dev = irq_data_get_irq_chip_data(d);
+
+ if (!its_dev->event_map.vm || !irqd_is_forwarded_to_vcpu(d))
+ return -EINVAL;
+
+ if (info->cmd_type == PROP_UPDATE_AND_INV_VLPI)
+ lpi_update_config(d, 0xff, info->config);
+ else
+ lpi_write_config(d, 0xff, info->config);
+ its_vlpi_set_doorbell(d, !!(info->config & LPI_PROP_ENABLED));
+
+ return 0;
+}
+
static int its_irq_set_vcpu_affinity(struct irq_data *d, void *vcpu_info)
{
struct its_device *its_dev = irq_data_get_irq_chip_data(d);
@@ -1039,6 +1103,7 @@ static int its_irq_set_vcpu_affinity(struct irq_data *d, void *vcpu_info)
case PROP_UPDATE_VLPI:
case PROP_UPDATE_AND_INV_VLPI:
+ return its_vlpi_prop_update(d, info);
default:
return -EINVAL;
--
2.11.0
next prev parent reply other threads:[~2017-06-28 15:17 UTC|newest]
Thread overview: 57+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-06-28 15:03 [PATCH v2 00/52] irqchip: KVM: Add support for GICv4 Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 01/52] genirq: Let irq_set_vcpu_affinity() iterate over hierarchy Marc Zyngier
2017-07-04 21:15 ` Thomas Gleixner
2017-06-28 15:03 ` [PATCH v2 02/52] irqchip/gic-v3: Add redistributor iterator Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 03/52] irqchip/gic-v3: Add VLPI/DirectLPI discovery Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 04/52] irqchip/gic-v3-its: Move LPI definitions around Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 05/52] irqchip/gic-v3-its: Add probing for VLPI properties Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 06/52] irqchip/gic-v3-its: Macro-ize its_send_single_command Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 07/52] irqchip/gic-v3-its: Implement irq_set_irqchip_state for pending state Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 08/52] irqchip/gic-v3-its: Split out property table allocation Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 09/52] irqchip/gic-v3-its: Allow use of indirect VCPU tables Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 10/52] irqchip/gic-v3-its: Split out pending table allocation Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 11/52] irqchip/gic-v3-its: Rework LPI freeing Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 12/52] irqchip/gic-v3-its: Generalize device table allocation Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 13/52] irqchip/gic-v3-its: Generalize LPI configuration Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 14/52] irqchip/gic-v4: Add management structure definitions Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 15/52] irqchip/gic-v3-its: Add GICv4 ITS command definitions Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 16/52] irqchip/gic-v3-its: Add VLPI configuration hook Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 17/52] irqchip/gic-v3-its: Add VLPI map/unmap operations Marc Zyngier
2017-06-28 15:03 ` Marc Zyngier [this message]
2017-06-28 15:03 ` [PATCH v2 19/52] irqchip/gic-v3-its: Add VPE domain infrastructure Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 20/52] irqchip/gic-v3-its: Add VPE irq domain allocation/teardown Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 21/52] irqchip/gic-v3-its: Add VPE irq domain [de]activation Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 22/52] irqchip/gic-v3-its: Add VPENDBASER/VPROPBASER accessors Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 23/52] irqchip/gic-v3-its: Add VPE scheduling Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 24/52] irqchip/gic-v3-its: Add VPE invalidation hook Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 25/52] irqchip/gic-v3-its: Add VPE affinity changes Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 26/52] irqchip/gic-v3-its: Add VPE interrupt masking Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 27/52] irqchip/gic-v3-its: Support VPE doorbell invalidation even when !DirectLPI Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 28/52] irqchip/gic-v3-its: Set implementation defined bit to enable VLPIs Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 29/52] irqchip/gic-v4: Add per-VM VPE domain creation Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 30/52] irqchip/gic-v4: Add VPE command interface Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 31/52] irqchip/gic-v4: Add VLPI configuration interface Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 32/52] irqchip/gic-v4: Add some basic documentation Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 33/52] irqchip/gic-v4: Enable low-level GICv4 operations Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 34/52] irqchip/gic-v3: Advertise GICv4 support to KVM Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 35/52] KVM: arm/arm64: vgic: Move kvm_vgic_destroy call around Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 36/52] KVM: arm/arm64: vITS: Add MSI translation helpers Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 37/52] KVM: arm/arm64: GICv4: Add init and teardown of the vPE irq domain Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 38/52] KVM: arm/arm64: GICv4: Wire init/teardown of per-VM support Marc Zyngier
2017-07-08 11:26 ` Shanker Donthineni
2017-07-10 16:34 ` Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 39/52] KVM: arm/arm64: GICv4: Wire mapping/unmapping of VLPIs in VFIO irq bypass Marc Zyngier
2017-06-28 15:03 ` [PATCH v2 40/52] KVM: arm/arm64: GICv4: Handle INT command applied to a VLPI Marc Zyngier
2017-06-28 15:04 ` [PATCH v2 41/52] KVM: arm/arm64: GICv4: Unmap VLPI when freeing an LPI Marc Zyngier
2017-06-28 15:04 ` [PATCH v2 42/52] KVM: arm/arm64: GICv4: Handle MOVI applied to a VLPI Marc Zyngier
2017-06-28 15:04 ` [PATCH v2 43/52] KVM: arm/arm64: GICv4: Handle CLEAR " Marc Zyngier
2017-06-28 15:04 ` [PATCH v2 44/52] KVM: arm/arm64: GICv4: Handle MOVALL applied to a vPE Marc Zyngier
2017-06-28 15:04 ` [PATCH v2 45/52] KVM: arm/arm64: GICv4: Propagate property updates to VLPIs Marc Zyngier
2017-06-28 15:04 ` [PATCH v2 46/52] KVM: arm/arm64: GICv4: Handle INVALL applied to a vPE Marc Zyngier
2017-06-28 15:04 ` [PATCH v2 47/52] KVM: arm/arm64: GICv4: Propagate VLPI properties at map time Marc Zyngier
2017-06-28 15:04 ` [PATCH v2 48/52] KVM: arm/arm64: GICv4: Add doorbell interrupt handling Marc Zyngier
2017-06-28 15:04 ` [PATCH v2 49/52] KVM: arm/arm64: GICv4: Hook vPE scheduling into vgic flush/sync Marc Zyngier
2017-06-28 15:04 ` [PATCH v2 50/52] KVM: arm/arm64: GICv4: Enable virtual cpuif if VLPIs can be delivered Marc Zyngier
2017-06-28 15:04 ` [PATCH v2 51/52] KVM: arm/arm64: GICv4: Use pending_last as a scheduling hint Marc Zyngier
2017-06-28 15:04 ` [PATCH v2 52/52] KVM: arm/arm64: GICv4: Enable VLPI support Marc Zyngier
2017-07-01 14:54 ` [PATCH v2 00/52] irqchip: KVM: Add support for GICv4 Shanker Donthineni
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=20170628150411.15846-19-marc.zyngier@arm.com \
--to=marc.zyngier@arm.com \
--cc=christoffer.dall@linaro.org \
--cc=eric.auger@redhat.com \
--cc=jason@lakedaemon.net \
--cc=kvmarm@lists.cs.columbia.edu \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mark.rutland@arm.com \
--cc=shankerd@codeaurora.org \
--cc=tglx@linutronix.de \
/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;
as well as URLs for NNTP newsgroup(s).