Skip to contentSkip to conversation

pci related patches 11-21-2009

DISCUSSION Yinghai Lu 3 participants 11 messages Nov 21, 2009 – Dec 16, 2009

Subsystems (2)
lore
Share & tools
lore
author
#1 lore raw

x86 related patches 11-21-2009

please check several patches from
git://git.kernel.org/pub/scm/linux/kernel/git/yinghai/linux-2.6-yinghai.git for-tip

YH

#2 lore raw

[tip:x86/apic] x86: apic: Remove not needed #ifdef

Commit-ID:  e670761f12f4069d204f433bf547d9c679a4fd05
Gitweb:     http://git.kernel.org/tip/e670761f12f4069d204f433bf547d9c679a4fd05
Author:     Yinghai Lu <[email protected]>
AuthorDate: Sat, 21 Nov 2009 00:23:37 -0800
Committer:  Ingo Molnar <[email protected]>
CommitDate: Mon, 23 Nov 2009 09:54:15 +0100

x86: apic: Remove not needed #ifdef

Suresh made dmar_table_init() already have that protection.

LKML-Reference: <[email protected]>

---
 arch/x86/kernel/apic/apic.c |    9 +++------
 1 files changed, 3 insertions(+), 6 deletions(-)
arch/x86/kernel/apic/apic.c +3 -6
index 4c689f4..ad8c75b 100644
@@ -1377,14 +1377,11 @@ void __init enable_IR_x2apic(void)
 	unsigned long flags;
 	struct IO_APIC_route_entry **ioapic_entries = NULL;
 	int ret, x2apic_enabled = 0;
-	int dmar_table_init_ret = 0;
+	int dmar_table_init_ret;
 
-#ifdef CONFIG_INTR_REMAP
 	dmar_table_init_ret = dmar_table_init();
-	if (dmar_table_init_ret)
-		pr_debug("dmar_table_init() failed with %d:\n",
-				dmar_table_init_ret);
-#endif
+	if (dmar_table_init_ret && !x2apic_supported())
+		return;
 
 	ioapic_entries = alloc_ioapic_entries();
 	if (!ioapic_entries) {
#3 lore raw

[tip:x86/apic] x86: Re-get cfg_new in case reuse/move irq_desc

In reply to Yinghai Lu
Commit-ID:  37ef2a3029fde884808ff1b369677abc7dd9a79a
Gitweb:     http://git.kernel.org/tip/37ef2a3029fde884808ff1b369677abc7dd9a79a
Author:     Yinghai Lu <[email protected]>
AuthorDate: Sat, 21 Nov 2009 00:23:37 -0800
Committer:  Ingo Molnar <[email protected]>
CommitDate: Mon, 23 Nov 2009 09:56:05 +0100

x86: Re-get cfg_new in case reuse/move irq_desc

When irq_desc is moved, we need to make sure to use the right cfg_new.

LKML-Reference: <[email protected]>

---
 arch/x86/kernel/apic/io_apic.c |    1 +
 1 files changed, 1 insertions(+), 0 deletions(-)
arch/x86/kernel/apic/io_apic.c +1 -0
index ff23719..085e60e 100644
@@ -3186,6 +3186,7 @@ unsigned int create_irq_nr(unsigned int irq_want, int node)
 			continue;
 
 		desc_new = move_irq_desc(desc_new, node);
+		cfg_new = desc_new->chip_data;
 
 		if (__assign_irq_vector(new, cfg_new, apic->target_cpus()) == 0)
 			irq = new;
#4 lore raw

[tip:x86/apic] x86: apic: Print out SRAT table APIC id in hex

In reply to Yinghai Lu
Commit-ID:  163d3866cfa79aa5945f1ee5e43fb3ed1455f75c
Gitweb:     http://git.kernel.org/tip/163d3866cfa79aa5945f1ee5e43fb3ed1455f75c
Author:     Yinghai Lu <[email protected]>
AuthorDate: Sat, 21 Nov 2009 00:23:37 -0800
Committer:  Ingo Molnar <[email protected]>
CommitDate: Mon, 23 Nov 2009 09:56:30 +0100

x86: apic: Print out SRAT table APIC id in hex

Make it consistent with APIC MADT print out, for big systems APIC id in hex is more readable.

LKML-Reference: <[email protected]>

---
 arch/x86/mm/srat_64.c |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)
arch/x86/mm/srat_64.c +2 -2
index dbb5381..9d7ce96 100644
@@ -136,7 +136,7 @@ acpi_numa_x2apic_affinity_init(struct acpi_srat_x2apic_cpu_affinity *pa)
 	apicid_to_node[apic_id] = node;
 	node_set(node, cpu_nodes_parsed);
 	acpi_numa = 1;
-	printk(KERN_INFO "SRAT: PXM %u -> APIC %u -> Node %u\n",
+	printk(KERN_INFO "SRAT: PXM %u -> APIC 0x%04x -> Node %u\n",
 	       pxm, apic_id, node);
 }
 
@@ -170,7 +170,7 @@ acpi_numa_processor_affinity_init(struct acpi_srat_cpu_affinity *pa)
 	apicid_to_node[apic_id] = node;
 	node_set(node, cpu_nodes_parsed);
 	acpi_numa = 1;
-	printk(KERN_INFO "SRAT: PXM %u -> APIC %u -> Node %u\n",
+	printk(KERN_INFO "SRAT: PXM %u -> APIC 0x%02x -> Node %u\n",
 	       pxm, apic_id, node);
 }
 
#5 lore raw

[tip:x86/mm] x86, numa, bootmem: Only free bootmem on NUMA failure path

In reply to Yinghai Lu
Commit-ID:  021428ad1418cf3c386a1a0157140c3ea29b17ef
Gitweb:     http://git.kernel.org/tip/021428ad1418cf3c386a1a0157140c3ea29b17ef
Author:     Yinghai Lu <[email protected]>
AuthorDate: Sat, 21 Nov 2009 00:23:37 -0800
Committer:  Ingo Molnar <[email protected]>
CommitDate: Mon, 23 Nov 2009 10:00:48 +0100

x86, numa, bootmem: Only free bootmem on NUMA failure path

In the NUMA bootmem setup failure path we freed nodedata_phys incorrectly.

Cc Thomas Gleixner <[email protected]>
Cc H. Peter Anvin <[email protected]>
Cc Rusty Russell <[email protected]>
Cc David Rientjes <[email protected]>
Cc Andrew Morton <[email protected]>

LKML-Reference: <[email protected]>

---
 arch/x86/mm/numa_64.c |   10 ++++++++--
 1 files changed, 8 insertions(+), 2 deletions(-)
arch/x86/mm/numa_64.c +8 -2
index 086f98a..3acd870 100644
@@ -239,8 +239,14 @@ setup_node_bootmem(int nodeid, unsigned long start, unsigned long end)
 	bootmap = early_node_mem(nodeid, bootmap_start, end,
 				 bootmap_pages<<PAGE_SHIFT, PAGE_SIZE);
 	if (bootmap == NULL)  {
-		if (nodedata_phys < start || nodedata_phys >= end)
-			free_bootmem(nodedata_phys, pgdat_size);
+		if (nodedata_phys < start || nodedata_phys >= end) {
+			/*
+			 * only need to free it if it is from other node
+			 * bootmem
+			 */
+			if (nid != nodeid)
+				free_bootmem(nodedata_phys, pgdat_size);
+		}
 		node_data[nodeid] = NULL;
 		return;
 	}
#6 lore raw

[tip:x86/mm] x86, numa: Use near(er) online node instead of roundrobin for NUMA

In reply to Yinghai Lu
Commit-ID:  d9c2d5ac6af87b4491bff107113aaf16f6c2b2d9
Gitweb:     http://git.kernel.org/tip/d9c2d5ac6af87b4491bff107113aaf16f6c2b2d9
Author:     Yinghai Lu <[email protected]>
AuthorDate: Sat, 21 Nov 2009 00:23:37 -0800
Committer:  Ingo Molnar <[email protected]>
CommitDate: Mon, 23 Nov 2009 10:06:24 +0100

x86, numa: Use near(er) online node instead of roundrobin for NUMA

CPU to node mapping is set via the following sequence:

1. numa_init_array(): Set up roundrobin from cpu to online node

2. init_cpu_to_node(): Set that according to apicid_to_node[] according to srat only handle the node that is online, and leave other cpu on node without ram (aka not online) to still roundrobin.

3. later call srat_detect_node for Intel/AMD, will use first_online node or nearby node.

Problem is that setup_per_cpu_areas() is not called between 2 and 3, the per_cpu for cpu on node with ram is on different node, and could put that on node with two hops away.

So try to optimize this and add find_near_online_node() and call init_cpu_to_node().

Cc Tejun Heo <[email protected]>
Cc Linus Torvalds <[email protected]>
Cc Thomas Gleixner <[email protected]>
Cc H. Peter Anvin <[email protected]>
Cc Rusty Russell <[email protected]>
Cc David Rientjes <[email protected]>
Cc Andrew Morton <[email protected]>

LKML-Reference: <[email protected]>

---
 arch/x86/kernel/cpu/intel.c |    6 +++++-
 arch/x86/mm/numa_64.c       |   21 ++++++++++++++++++++-
 2 files changed, 25 insertions(+), 2 deletions(-)
arch/x86/kernel/cpu/intel.c +5 -1
index 40e1835..c900b73 100644
@@ -263,8 +263,12 @@ static void __cpuinit srat_detect_node(struct cpuinfo_x86 *c)
 	/* Don't do the funky fallback heuristics the AMD version employs
 	   for now. */
 	node = apicid_to_node[apicid];
-	if (node == NUMA_NO_NODE || !node_online(node))
+	if (node == NUMA_NO_NODE)
 		node = first_node(node_online_map);
+	else if (!node_online(node)) {
+		/* reuse the value from init_cpu_to_node() */
+		node = cpu_to_node(cpu);
+	}
 	numa_set_node(cpu, node);
 
 	printk(KERN_INFO "CPU %d/0x%x -> Node %d\n", cpu, apicid, node);
arch/x86/mm/numa_64.c +20 -1
index 3acd870..83bbc70 100644
@@ -764,6 +764,25 @@ static __init int numa_setup(char *opt)
 early_param("numa", numa_setup);
 
 #ifdef CONFIG_NUMA
+
+static __init int find_near_online_node(int node)
+{
+	int n, val;
+	int min_val = INT_MAX;
+	int best_node = -1;
+
+	for_each_online_node(n) {
+		val = node_distance(node, n);
+
+		if (val < min_val) {
+			min_val = val;
+			best_node = n;
+		}
+	}
+
+	return best_node;
+}
+
 /*
  * Setup early cpu_to_node.
  *
@@ -795,7 +814,7 @@ void __init init_cpu_to_node(void)
 		if (node == NUMA_NO_NODE)
 			continue;
 		if (!node_online(node))
-			continue;
+			node = find_near_online_node(node);
 		numa_set_node(cpu, node);
 	}
 }
#8 lore raw
In reply to Yinghai Lu
Quoted 6 lines · On Sat, 21 Nov 2009 00:35:36 -0800 Yinghai Lu <[email protected]> wrote:

please check several patches from
git://git.kernel.org/pub/scm/linux/kernel/git/yinghai/linux-2.6-yinghai.git
for-pci

http://git.kernel.org/?p=linux/kernel/git/yinghai/linux-2.6-yinghai.git;a=shortlog;h=refs/heads/for-pci

The DMAR stuff can just go through the DMAR tree.

I don't see the two debug output patches on linux-pci, did you post them there?

I commented on one of the resource patches, I'll check the other two (long standing leaf release and pciehp) ones again; I mainly want a bunch of tested-bys for those.

The x86_pci_rootbus one looks ok I think, I'll see about applying that today.

Signature · 1 line
Jesse Barnes, Intel Open Source Technology Center
author
#9 lore raw
Quoted 11 lines · Jesse Barnes
Quoted 6 lines · On Sat, 21 Nov 2009 00:35:36 -0800 Yinghai Lu <[email protected]> wrote:

please check several patches from
git://git.kernel.org/pub/scm/linux/kernel/git/yinghai/linux-2.6-yinghai.git
for-pci

http://git.kernel.org/?p=linux/kernel/git/yinghai/linux-2.6-yinghai.git;a=shortlog;h=refs/heads/for-pci

The DMAR stuff can just go through the DMAR tree.

sent to David Woodhouse two times.

thinking you could pick up too. it is under drivers/pci ?

Quoted 3 lines

I don't see the two debug output patches on linux-pci, did you post them there?

will post them there.

Quoted 4 lines

I commented on one of the resource patches, I'll check the other two (long standing leaf release and pciehp) ones again; I mainly want a bunch of tested-bys for those.

thanks. thought will get one Reviewed-by and Tested-by from Kenji. still need to address bridges resource shrink hotplug bridges other than pcie. already sent him one add-on patch.

Quoted 3 lines

The x86_pci_rootbus one looks ok I think, I'll see about applying that today.

Thanks

Yinghai

#10 lore raw

* Yinghai Lu <[email protected]>:

Quoted 9 lines
Quoted 4 lines · Jesse Barnes

I commented on one of the resource patches, I'll check the other two (long standing leaf release and pciehp) ones again; I mainly want a bunch of tested-bys for those.

thanks. thought will get one Reviewed-by and Tested-by from Kenji. still need to address bridges resource shrink hotplug bridges other than pcie. already sent him one add-on patch.

Sorry, I have been swamped lately with other things.

I will try and test Yinghai's patches this week.

/ac

author
#7 lore raw

x86 related patches 12-15-2009

In reply to Yinghai Lu

please check several patches from
git://git.kernel.org/pub/scm/linux/kernel/git/yinghai/linux-2.6-yinghai.git for-tip

http://git.kernel.org/?p=linux/kernel/git/yinghai/linux-2.6-yinghai.git;a=shortlog;h=refs/heads/for-tip

1a88d3a: x86: check range in update range
0fb059e: x86: move range related operation to one file
8a81618: x86: only call dma32_reserve_bootmem 64bit !CONFIG_NUMA 75d6a28: x86: make early_node_mem get mem > 4g if possible -v2 557bcb7: x86: print bootmem free before pci_iommu_alloc and free_all_bootmem 72fe68a: x86: dynamic increase early_res array size
de4546b: x86: introduce max_early_res and early_res_count 1caafab: x86: call early_res_to_bootmem one time
daefc77: x86: increase MAX_EARLY_RES
a9dc1ac: x86: fix checking of SRAT when node0 ram is not from 0 -v3

Thanks

Yinghai