Commit 80cce3cd authored by Ganapatrao Kulkarni's avatar Ganapatrao Kulkarni Committed by Greg Kroah-Hartman

irqchip/gicv3-its: numa: Enable workaround for Cavium thunderx erratum 23144

[ Upstream commit fbf8f40e ]

The erratum fixes the hang of ITS SYNC command by avoiding inter node
io and collections/cpu mapping on thunderx dual-socket platform.

This fix is only applicable for Cavium's ThunderX dual-socket platform.
Reviewed-by: default avatarRobert Richter <rrichter@cavium.com>
Signed-off-by: default avatarGanapatrao Kulkarni <gkulkarni@caviumnetworks.com>
Signed-off-by: default avatarRobert Richter <rrichter@cavium.com>
Signed-off-by: default avatarMarc Zyngier <marc.zyngier@arm.com>
Signed-off-by: default avatarSasha Levin <alexander.levin@verizon.com>
Signed-off-by: default avatarGreg Kroah-Hartman <gregkh@linuxfoundation.org>
parent a2350f3d
...@@ -391,6 +391,15 @@ config CAVIUM_ERRATUM_22375 ...@@ -391,6 +391,15 @@ config CAVIUM_ERRATUM_22375
If unsure, say Y. If unsure, say Y.
config CAVIUM_ERRATUM_23144
bool "Cavium erratum 23144: ITS SYNC hang on dual socket system"
depends on NUMA
default y
help
ITS SYNC command hang for cross node io and collections/cpu mapping.
If unsure, say Y.
config CAVIUM_ERRATUM_23154 config CAVIUM_ERRATUM_23154
bool "Cavium erratum 23154: Access to ICC_IAR1_EL1 is not sync'ed" bool "Cavium erratum 23154: Access to ICC_IAR1_EL1 is not sync'ed"
default y default y
......
...@@ -41,6 +41,7 @@ ...@@ -41,6 +41,7 @@
#define ITS_FLAGS_CMDQ_NEEDS_FLUSHING (1ULL << 0) #define ITS_FLAGS_CMDQ_NEEDS_FLUSHING (1ULL << 0)
#define ITS_FLAGS_WORKAROUND_CAVIUM_22375 (1ULL << 1) #define ITS_FLAGS_WORKAROUND_CAVIUM_22375 (1ULL << 1)
#define ITS_FLAGS_WORKAROUND_CAVIUM_23144 (1ULL << 2)
#define RDIST_FLAGS_PROPBASE_NEEDS_FLUSHING (1 << 0) #define RDIST_FLAGS_PROPBASE_NEEDS_FLUSHING (1 << 0)
...@@ -71,6 +72,7 @@ struct its_node { ...@@ -71,6 +72,7 @@ struct its_node {
struct list_head its_device_list; struct list_head its_device_list;
u64 flags; u64 flags;
u32 ite_size; u32 ite_size;
int numa_node;
}; };
#define ITS_ITT_ALIGN SZ_256 #define ITS_ITT_ALIGN SZ_256
...@@ -600,11 +602,23 @@ static void its_unmask_irq(struct irq_data *d) ...@@ -600,11 +602,23 @@ static void its_unmask_irq(struct irq_data *d)
static int its_set_affinity(struct irq_data *d, const struct cpumask *mask_val, static int its_set_affinity(struct irq_data *d, const struct cpumask *mask_val,
bool force) bool force)
{ {
unsigned int cpu = cpumask_any_and(mask_val, cpu_online_mask); unsigned int cpu;
const struct cpumask *cpu_mask = cpu_online_mask;
struct its_device *its_dev = irq_data_get_irq_chip_data(d); struct its_device *its_dev = irq_data_get_irq_chip_data(d);
struct its_collection *target_col; struct its_collection *target_col;
u32 id = its_get_event_id(d); u32 id = its_get_event_id(d);
/* 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) {
cpu_mask = cpumask_of_node(its_dev->its->numa_node);
if (!cpumask_intersects(mask_val, cpu_mask))
return -EINVAL;
}
}
cpu = cpumask_any_and(mask_val, cpu_mask);
if (cpu >= nr_cpu_ids) if (cpu >= nr_cpu_ids)
return -EINVAL; return -EINVAL;
...@@ -1081,6 +1095,16 @@ static void its_cpu_init_collection(void) ...@@ -1081,6 +1095,16 @@ static void its_cpu_init_collection(void)
list_for_each_entry(its, &its_nodes, entry) { list_for_each_entry(its, &its_nodes, entry) {
u64 target; u64 target;
/* avoid cross node collections and its mapping */
if (its->flags & ITS_FLAGS_WORKAROUND_CAVIUM_23144) {
struct device_node *cpu_node;
cpu_node = of_get_cpu_node(cpu, NULL);
if (its->numa_node != NUMA_NO_NODE &&
its->numa_node != of_node_to_nid(cpu_node))
continue;
}
/* /*
* We now have to bind each collection to its target * We now have to bind each collection to its target
* redistributor. * redistributor.
...@@ -1308,9 +1332,14 @@ static void its_irq_domain_activate(struct irq_domain *domain, ...@@ -1308,9 +1332,14 @@ static void its_irq_domain_activate(struct irq_domain *domain,
{ {
struct its_device *its_dev = irq_data_get_irq_chip_data(d); struct its_device *its_dev = irq_data_get_irq_chip_data(d);
u32 event = its_get_event_id(d); u32 event = its_get_event_id(d);
const struct cpumask *cpu_mask = cpu_online_mask;
/* get the cpu_mask of local node */
if (its_dev->its->numa_node >= 0)
cpu_mask = cpumask_of_node(its_dev->its->numa_node);
/* Bind the LPI to the first possible CPU */ /* Bind the LPI to the first possible CPU */
its_dev->event_map.col_map[event] = cpumask_first(cpu_online_mask); its_dev->event_map.col_map[event] = cpumask_first(cpu_mask);
/* Map the GIC IRQ and event to the device */ /* Map the GIC IRQ and event to the device */
its_send_mapvi(its_dev, d->hwirq, event); its_send_mapvi(its_dev, d->hwirq, event);
...@@ -1400,6 +1429,13 @@ static void __maybe_unused its_enable_quirk_cavium_22375(void *data) ...@@ -1400,6 +1429,13 @@ static void __maybe_unused its_enable_quirk_cavium_22375(void *data)
its->flags |= ITS_FLAGS_WORKAROUND_CAVIUM_22375; its->flags |= ITS_FLAGS_WORKAROUND_CAVIUM_22375;
} }
static void __maybe_unused its_enable_quirk_cavium_23144(void *data)
{
struct its_node *its = data;
its->flags |= ITS_FLAGS_WORKAROUND_CAVIUM_23144;
}
static const struct gic_quirk its_quirks[] = { static const struct gic_quirk its_quirks[] = {
#ifdef CONFIG_CAVIUM_ERRATUM_22375 #ifdef CONFIG_CAVIUM_ERRATUM_22375
{ {
...@@ -1408,6 +1444,14 @@ static const struct gic_quirk its_quirks[] = { ...@@ -1408,6 +1444,14 @@ static const struct gic_quirk its_quirks[] = {
.mask = 0xffff0fff, .mask = 0xffff0fff,
.init = its_enable_quirk_cavium_22375, .init = its_enable_quirk_cavium_22375,
}, },
#endif
#ifdef CONFIG_CAVIUM_ERRATUM_23144
{
.desc = "ITS: Cavium erratum 23144",
.iidr = 0xa100034c, /* ThunderX pass 1.x */
.mask = 0xffff0fff,
.init = its_enable_quirk_cavium_23144,
},
#endif #endif
{ {
} }
...@@ -1470,6 +1514,7 @@ static int its_probe(struct device_node *node, struct irq_domain *parent) ...@@ -1470,6 +1514,7 @@ static int its_probe(struct device_node *node, struct irq_domain *parent)
its->base = its_base; its->base = its_base;
its->phys_base = res.start; its->phys_base = res.start;
its->ite_size = ((readl_relaxed(its_base + GITS_TYPER) >> 4) & 0xf) + 1; its->ite_size = ((readl_relaxed(its_base + GITS_TYPER) >> 4) & 0xf) + 1;
its->numa_node = of_node_to_nid(node);
its->cmd_base = kzalloc(ITS_CMD_QUEUE_SZ, GFP_KERNEL); its->cmd_base = kzalloc(ITS_CMD_QUEUE_SZ, GFP_KERNEL);
if (!its->cmd_base) { if (!its->cmd_base) {
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment