Re: [RFC] Remove unswappable anonymous pages off the LRU

[Martin Bligh <mbligh@mbligh.org>]

>>> #define PageMlocked(page)	(page->lru.next == some_address_which_isnt_used_for_anwything_else)
>> Mine just created a locked list. If you stick them there, there's no
>> need for a page flag ... and we don't abuse the lru pointers AGAIN! ;-)
> 
> I don't think there's a need for a mlocked list in the mlock patches:
> nothing ever needs to walk it.
> 
> However this might be a good way of solving the someone-did-a-swapon
> problem for this anon patch.
> 
> Guys, this page-flag problem is really serious.  -mm adds PG_mlocked and
> PG_readahead and the ext4 patches add PG_booked (am currently fighting the
> good fight there).  There's ongoing steady growth in these things and soon
> we're going to be in a lot of pain.

Well, if the list is sufficient to fix that, I don't see why we'd
care about the overhead of list manipulation vs a flag, it's not
a fast path.

>> Suspect most of the rest of my patch is crap, but that might be useful?
> 
> wordwrapped, space-stuffed and tab-replaced.  The trifecta!

That's cause it was fairly obviously useless as-was so I just cut
and pasted it. But nonetheless, I appreciate your adulation ;-)

I'll try to add CamelCaps, bracing fuckups, and lots of #ifdefs
for the next round.

M.

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>