From: Nathan Lynch <nathanl@linux.ibm.com>
To: Laurent Dufour <ldufour@linux.ibm.com>
Cc: cheloha@linux.ibm.com, linux-kernel@vger.kernel.org,
paulus@samba.org, linuxppc-dev@lists.ozlabs.org
Subject: Re: [PATCH v3] pseries: prevent free CPU ids to be reused on another node
Date: Wed, 07 Apr 2021 09:55:37 -0500 [thread overview]
Message-ID: <87blaqkuty.fsf@linux.ibm.com> (raw)
In-Reply-To: <20210406182554.85197-1-ldufour@linux.ibm.com>
Laurent Dufour <ldufour@linux.ibm.com> writes:
> Changes since V2, addressing Nathan's comments:
> - Remove the retry feature
> - Reduce the number of local variables (removing 'i')
I was more interested in not having two variables for NUMA nodes in the
function named 'node' and 'nid', hoping at least one of them could have
a more descriptive name. See below.
> static int pseries_add_processor(struct device_node *np)
> {
> - unsigned int cpu;
> + unsigned int cpu, node;
> cpumask_var_t candidate_mask, tmp;
> - int err = -ENOSPC, len, nthreads, i;
> + int err = -ENOSPC, len, nthreads, nid;
> const __be32 *intserv;
>
> intserv = of_get_property(np, "ibm,ppc-interrupt-server#s", &len);
> @@ -163,9 +169,17 @@ static int pseries_add_processor(struct device_node *np)
> zalloc_cpumask_var(&candidate_mask, GFP_KERNEL);
> zalloc_cpumask_var(&tmp, GFP_KERNEL);
>
> + /*
> + * Fetch from the DT nodes read by dlpar_configure_connector() the NUMA
> + * node id the added CPU belongs to.
> + */
> + nid = of_node_to_nid(np);
> + if (nid < 0 || !node_possible(nid))
> + nid = first_online_node;
> +
> nthreads = len / sizeof(u32);
> - for (i = 0; i < nthreads; i++)
> - cpumask_set_cpu(i, tmp);
> + for (cpu = 0; cpu < nthreads; cpu++)
> + cpumask_set_cpu(cpu, tmp);
>
> cpu_maps_update_begin();
>
> @@ -173,6 +187,19 @@ static int pseries_add_processor(struct device_node *np)
>
> /* Get a bitmap of unoccupied slots. */
> cpumask_xor(candidate_mask, cpu_possible_mask, cpu_present_mask);
> +
> + /*
> + * Remove free ids previously assigned on the other nodes. We can walk
> + * only online nodes because once a node became online it is not turned
> + * offlined back.
> + */
> + for_each_online_node(node) {
> + if (node == nid) /* Keep our node's recorded ids */
> + continue;
> + cpumask_andnot(candidate_mask, candidate_mask,
> + node_recorded_ids_map[node]);
> + }
> +
e.g. change 'nid' to 'assigned_node' or similar, and I think this
becomes easier to follow.
Otherwise the patch looks fine to me now.
next prev parent reply other threads:[~2021-04-07 14:56 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-04-06 18:25 [PATCH v3] pseries: prevent free CPU ids to be reused on another node Laurent Dufour
2021-04-07 14:55 ` Nathan Lynch [this message]
2021-04-07 15:00 ` Laurent Dufour
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=87blaqkuty.fsf@linux.ibm.com \
--to=nathanl@linux.ibm.com \
--cc=cheloha@linux.ibm.com \
--cc=ldufour@linux.ibm.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=paulus@samba.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;
as well as URLs for NNTP newsgroup(s).