All of lore.kernel.org
 help / color / mirror / Atom feed
From: Noam Camus <noamca@mellanox.com>
To: <decot@googlers.com>
Cc: <ben@decadent.org.uk>, <davem@davemloft.net>,
	<akpm@linux-foundation.org>, <xinhuix.pan@intel.com>,
	<linux-kernel@vger.kernel.org>, Noam Camus <noamca@mellanox.com>
Subject: [PATCH] lib/bitmap.c: enhance map pattern
Date: Sat, 16 Jul 2016 10:57:23 +0300	[thread overview]
Message-ID: <1468655843-13697-1-git-send-email-noamca@mellanox.com> (raw)

From: Noam Camus <noamca@mellanox.com>

Today there are platforms with many CPUs (up to 4K).
Trying to boot only part of the CPUs may result in too long string.

For example lets take NPS platform that is part of arch/arc.
This platform have SMP system with 256 cores each with
16 HW threads (SMT machine) where HW thread appears as CPU to the kernel.
In this example there is total of 4K CPUs.
When one tries to boot only part of the HW threads from each core the
string representing the map may be long...
For example if for sake of performance we decided to boot only first half
of HW threads of each core the map will look like:
0-7,16-23,32,39,...,4080-4087

This patch introduce new format to accommodate with such use case.
I added an optional postfix to a range of CPUs which will choose
according to given modulo the desired range of reminders i.e.:
<cpu range>%modulo=<range of reminders>

For example above pattern can be described in new format like this:
0-4095%16=0-7

Note that this patch is backward compatible.

Signed-off-by: Noam Camus <noamca@mellanox.com>
---
 lib/bitmap.c |   44 ++++++++++++++++++++++++++++++++++++++++----
 1 files changed, 40 insertions(+), 4 deletions(-)

diff --git a/lib/bitmap.c b/lib/bitmap.c
index c66da50..4f446bc 100644
--- a/lib/bitmap.c
+++ b/lib/bitmap.c
@@ -496,6 +496,9 @@ EXPORT_SYMBOL(bitmap_print_to_pagebuf);
  * ranges.  Consecutively set bits are shown as two hyphen-separated
  * decimal numbers, the smallest and largest bit numbers set in
  * the range.
+ * Optionally each range can be postfixed to denote that only parts of it
+ * should be set. The parts are the range of reminders modulo some value.
+ * i.e. range%mod=rem_range e.g. 0-1023%256=0-2 ==> 0,1,256,257,512,513,768,769
  *
  * Returns 0 on success, -errno on invalid input strings.
  * Error values:
@@ -507,12 +510,14 @@ static int __bitmap_parselist(const char *buf, unsigned int buflen,
 		int is_user, unsigned long *maskp,
 		int nmaskbits)
 {
-	unsigned a, b;
+	unsigned int a, b, old_a, old_b, mod_val, mod_a, mod_b;
 	int c, old_c, totaldigits, ndigits;
 	const char __user __force *ubuf = (const char __user __force *)buf;
 	int at_start, in_range;
 
 	totaldigits = c = 0;
+	old_a = old_b = 0;
+	mod_val = mod_a = mod_b = 0;
 	bitmap_zero(maskp, nmaskbits);
 	do {
 		at_start = 1;
@@ -547,6 +552,23 @@ static int __bitmap_parselist(const char *buf, unsigned int buflen,
 			if ((totaldigits != ndigits) && isspace(old_c))
 				return -EINVAL;
 
+			if (c == '=') {
+				mod_val = a;
+				at_start = 1;
+				in_range = 0;
+				a = b = 0;
+				continue;
+			}
+
+			if (c == '%') {
+				old_a = a;
+				old_b = b;
+				at_start = 1;
+				in_range = 0;
+				a = b = 0;
+				continue;
+			}
+
 			if (c == '-') {
 				if (at_start || in_range)
 					return -EINVAL;
@@ -567,17 +589,31 @@ static int __bitmap_parselist(const char *buf, unsigned int buflen,
 		}
 		if (ndigits == totaldigits)
 			continue;
+		if (mod_val) {
+			mod_a = a;
+			mod_b = b;
+			a = old_a;
+			b = old_b;
+			old_a = old_b = 0;
+		}
 		/* if no digit is after '-', it's wrong*/
 		if (at_start && in_range)
 			return -EINVAL;
-		if (!(a <= b))
+		if (!(a <= b) || !(mod_a <= mod_b))
 			return -EINVAL;
-		if (b >= nmaskbits)
+		if (b >= nmaskbits || (mod_val && (mod_b >= mod_val)))
 			return -ERANGE;
 		while (a <= b) {
-			set_bit(a, maskp);
+			if (mod_val) {
+				unsigned int rem = a % mod_val;
+
+				if (rem >= mod_a && rem <= mod_b)
+					set_bit(a, maskp);
+			} else
+				set_bit(a, maskp);
 			a++;
 		}
+		mod_val = mod_a = mod_b = 0;
 	} while (buflen && c == ',');
 	return 0;
 }
-- 
1.7.1

             reply	other threads:[~2016-07-16 21:30 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-07-16  7:57 Noam Camus [this message]
2016-07-16 12:09 ` [PATCH] lib/bitmap.c: enhance map pattern Ben Hutchings
2016-07-17  6:33 ` [PATCH v2] lib/bitmap.c: enhance bitmap syntax Noam Camus
2016-07-17  8:10   ` [PATCH v3] " Noam Camus
2016-07-17 13:50     ` Pan Xinhui
2016-07-17 13:59       ` Pan Xinhui
2016-07-18  7:06         ` Noam Camus
2016-08-01  4:53           ` Noam Camus
2016-09-06  9:12           ` Noam Camus
2016-09-06 13:36             ` Ben Hutchings
2016-09-08 15:33     ` [PATCH v4] " Noam Camus
2016-09-08 22:19       ` Andrew Morton
2016-09-11  7:40       ` [PATCH v5] " Noam Camus
2016-09-13 22:54         ` Andrew Morton
2016-09-14  6:52           ` Noam Camus

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=1468655843-13697-1-git-send-email-noamca@mellanox.com \
    --to=noamca@mellanox.com \
    --cc=akpm@linux-foundation.org \
    --cc=ben@decadent.org.uk \
    --cc=davem@davemloft.net \
    --cc=decot@googlers.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=xinhuix.pan@intel.com \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.