From: Alex Williamson <alex.williamson@hp.com>
To: linux-ia64@vger.kernel.org
Subject: [PATCH] fixup pxm_to_nid_map
Date: Tue, 31 Aug 2004 21:38:18 +0000 [thread overview]
Message-ID: <1093988299.4620.37.camel@tdi> (raw)
I'd like to be able to use pxm_to_nid_map in several places to
discover proximity domain to node id associations. Unfortunately, after
reassign_cpu_only_nodes() plays with the node space, the pxm_to_nid_map
doesn't necessarily reflect reality. This fixes up the table so it's
still valid. Note that nid_to_pxm_map is still potentially broken, but
has a one-to-many problem if the above function combines several
proximity domains into a single node. Thanks to Bob Picco for the base
patch.
Signed-off-by: Alex Williamson <alex.williamson@hp.com>
=== arch/ia64/mm/discontig.c 1.18 vs edited ==--- 1.18/arch/ia64/mm/discontig.c 2004-08-27 00:30:29 -06:00
+++ edited/arch/ia64/mm/discontig.c 2004-08-31 14:44:45 -06:00
@@ -53,11 +53,12 @@
static void __init reassign_cpu_only_nodes(void)
{
struct node_memblk_s *p;
- int i, j, k, nnode, nid, cpu, cpunid;
+ int i, j, k, nnode, nid, cpu, cpunid, pxm;
u8 cslit, slit;
static DECLARE_BITMAP(nodes_with_mem, NR_NODES) __initdata;
static u8 numa_slit_fix[MAX_NUMNODES * MAX_NUMNODES] __initdata;
static int node_flip[NR_NODES] __initdata;
+ static int old_nid_map[NR_CPUS] __initdata;
for (nnode = 0, p = &node_memblk[0]; p < &node_memblk[num_node_memblks]; p++)
if (!test_bit(p->nid, (void *) nodes_with_mem)) {
@@ -104,9 +105,14 @@
for (cpu = 0; cpu < NR_CPUS; cpu++)
if (node_cpuid[cpu].nid = i) {
- /* For nodes not being reassigned just fix the cpu's nid. */
+ /*
+ * For nodes not being reassigned just
+ * fix the cpu's nid and reverse pxm map
+ */
if (cpunid < numnodes) {
- node_cpuid[cpu].nid = cpunid;
+ pxm = nid_to_pxm_map[i];
+ pxm_to_nid_map[pxm] + node_cpuid[cpu].nid = cpunid;
continue;
}
@@ -126,6 +132,8 @@
}
}
+ /* save old nid map so we can update the pxm */
+ old_nid_map[cpu] = node_cpuid[cpu].nid;
node_cpuid[cpu].nid = k;
}
}
@@ -134,14 +142,19 @@
* Fixup temporary nid values for CPU-only nodes.
*/
for (cpu = 0; cpu < NR_CPUS; cpu++)
- if (node_cpuid[cpu].nid = (numnodes + numnodes))
- node_cpuid[cpu].nid = nnode - 1;
- else
- for (i = 0; i < nnode; i++)
- if (node_flip[i] = (node_cpuid[cpu].nid - numnodes)) {
- node_cpuid[cpu].nid = i;
- break;
- }
+ if (node_cpuid[cpu].nid = (numnodes + numnodes)) {
+ pxm = nid_to_pxm_map[old_nid_map[cpu]];
+ pxm_to_nid_map[pxm] = node_cpuid[cpu].nid = nnode - 1;
+ } else {
+ for (i = 0; i < nnode; i++) {
+ if (node_flip[i] != (node_cpuid[cpu].nid - numnodes))
+ continue;
+
+ pxm = nid_to_pxm_map[old_nid_map[cpu]];
+ pxm_to_nid_map[pxm] = node_cpuid[cpu].nid = i;
+ break;
+ }
+ }
/*
* Fix numa_slit by compressing from larger
reply other threads:[~2004-08-31 21:38 UTC|newest]
Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
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=1093988299.4620.37.camel@tdi \
--to=alex.williamson@hp.com \
--cc=linux-ia64@vger.kernel.org \
/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