Skip to contentSkip to conversation

- Complete IRQ interrupt migration in arch_enable_uv_irq()

PATCH In review Jack Steiner 1 participant 2 messages Jul 20, 2009 – Aug 4, 2009 x86

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 - Complete IRQ interrupt migration in arch_enable_uv_irq()

In uv_setup_irq(), the call to create_irq() initially assigns IRQ vectors to cpu 0. The subsequent call to assign_irq_vector() in arch_enable_uv_irq() migrates the IRQ to another cpu and frees the cpu 0 vector - at least it will be freed as soon as the "IRQ move" completes.

arch_enable_uv_irq() needs to send a cleanup IPI to complete the IRQ move. Otherwise, assignment of GRU interrupts on large systems (>200 cpus) will exhaust the cpu 0 interrupt vectors and initialization of the GRU driver will fail.

---
 arch/x86/kernel/apic/io_apic.c |    3 +++
 1 file changed, 3 insertions(+)
arch/x86/kernel/apic/io_apic.c +3 -0
Index: linux/arch/x86/kernel/apic/io_apic.c
===================================================================
@@ -3737,6 +3737,9 @@ int arch_enable_uv_irq(char *irq_name, u
 	mmr_pnode = uv_blade_to_pnode(mmr_blade);
 	uv_write_global_mmr64(mmr_pnode, mmr_offset, mmr_value);
 
+	if (cfg->move_in_progress)
+		send_cleanup_vector(cfg);
+
 	return irq;
 }
 
#1 lore raw

[tip:x86/urgent] x86, UV: Complete IRQ interrupt migration in arch_enable_uv_irq()

Commit-ID:  2a5ef41661b56cf4eee042a6967c4e14b63e8eac
Gitweb:     http://git.kernel.org/tip/2a5ef41661b56cf4eee042a6967c4e14b63e8eac
Author:     Jack Steiner <[email protected]>
AuthorDate: Mon, 20 Jul 2009 09:28:41 -0500
Committer:  Ingo Molnar <[email protected]>
CommitDate: Tue, 4 Aug 2009 16:32:52 +0200

x86, UV: Complete IRQ interrupt migration in arch_enable_uv_irq()

In uv_setup_irq(), the call to create_irq() initially assigns IRQ vectors to cpu 0. The subsequent call to assign_irq_vector() in arch_enable_uv_irq() migrates the IRQ to another cpu and frees the cpu 0 vector - at least it will be freed as soon as the "IRQ move" completes.

arch_enable_uv_irq() needs to send a cleanup IPI to complete the IRQ move. Otherwise, assignment of GRU interrupts on large systems (>200 cpus) will exhaust the cpu 0 interrupt vectors and initialization of the GRU driver will fail.

LKML-Reference: <[email protected]>

---
 arch/x86/kernel/apic/io_apic.c |    3 +++
 1 files changed, 3 insertions(+), 0 deletions(-)
arch/x86/kernel/apic/io_apic.c +3 -0
index 2284a48..d2ed6c5 100644
@@ -3793,6 +3793,9 @@ int arch_enable_uv_irq(char *irq_name, unsigned int irq, int cpu, int mmr_blade,
 	mmr_pnode = uv_blade_to_pnode(mmr_blade);
 	uv_write_global_mmr64(mmr_pnode, mmr_offset, mmr_value);
 
+	if (cfg->move_in_progress)
+		send_cleanup_vector(cfg);
+
 	return irq;
 }