linux-kernel - Re: [PATCH 3/3] percpu: add macro PCPU_CHUNK_AREA_IN

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20150722002522.GB1834@dhcp-17-102.nay.redhat.com>
Date:	Wed, 22 Jul 2015 08:25:22 +0800
From:	Baoquan He <bhe@...hat.com>
To:	Christoph Lameter <cl@...ux.com>
Cc:	tj@...nel.org, linux-mm@...ck.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 3/3] percpu: add macro PCPU_CHUNK_AREA_IN_USE

Hi Christoph,

On 07/20/15 at 10:35am, Christoph Lameter wrote:
> On Mon, 20 Jul 2015, Baoquan He wrote:
> 
> > chunk->map[] contains <offset|in-use flag> of each area. Now add a
> > new macro PCPU_CHUNK_AREA_IN_USE and use it as the in-use flag to
> > replace all magic number '1'.
> 
> Hmmm... This is a bitflag and the code now looks like there is some sort
> of bitmask that were are using. Use bitops or something else that clearly
> implies that a bit is flipped instead?

Thanks for your reviewing and suggesting.

I tried your suggestion and changed to use set_bit/clear_bit to do
instead. It's like this:

@@ -328,8 +329,10 @@ static void pcpu_mem_free(void *ptr, size_t size)
  */
 static int pcpu_count_occupied_pages(struct pcpu_chunk *chunk, int i)
 {
-       int off = chunk->map[i] & ~1;
-       int end = chunk->map[i + 1] & ~1;
+       int off = chunk->map[i];
+       int end = chunk->map[i + 1];
+       clear_bit(PCPU_CHUNK_AREA_IN_USE_BIT, &chunk->map[i]);
+       clear_bit(PCPU_CHUNK_AREA_IN_USE_BIT, &chunk->map[i + 1]);

Looks like code becomes a little redundent. If several different bits in
chunk->map[] have different usage and need several different flags,
bitops maybe better. While now only the lowest bit need be handle, use
bitops kindof too much and can make code a little messy.

You and Tejun may be a little struggled on this change since it make
code longer. Tejun has suggested that at least use a shorter name, like
PCPU_MAP_BUSY. I am going to post v2 to see if it's better.

Thanks
Baoquan
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/