linux-kernel - Re: [PATCH 11/35] lmb: Add find_memory_core

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: <1273825187.21352.590.camel@pasglop>
Date:	Fri, 14 May 2010 18:19:47 +1000
From:	Benjamin Herrenschmidt <benh@...nel.crashing.org>
To:	Yinghai <yinghai.lu@...cle.com>
Cc:	Ingo Molnar <mingo@...e.hu>, Thomas Gleixner <tglx@...utronix.de>,
	"H. Peter Anvin" <hpa@...or.com>,
	Andrew Morton <akpm@...ux-foundation.org>,
	David Miller <davem@...emloft.net>,
	Linus Torvalds <torvalds@...ux-foundation.org>,
	Johannes Weiner <hannes@...xchg.org>,
	linux-kernel@...r.kernel.org, linux-arch@...r.kernel.org
Subject: Re: [PATCH 11/35] lmb: Add find_memory_core_early()

On Thu, 2010-05-13 at 23:39 -0700, Yinghai wrote:
> > Oh and this wont' work on sparc. You should probably instead add a
> > lmb_find_in_nid() to lmb which shares code with lmb_alloc_nid().
> 
> should work with sparc...
> 
> +               if (addr == LMB_ERROR)
> +                       continue;
> 
> should do the trick.

Ugh ?

OK, I'll have to let Davem deal with the fine point of the sparc bits,
but I think basically sparc has CONFIG_ARCH_POPULATES_NODE_MAP set, but
the way it's NUMA affinity works, the early_node_map[] is crap, you
cannot rely on the ranges in there.

So it will "work" in the sense that you won't get errors, but the
allocations will -not- be node local. At least that's my understanding.

There's a reason we are adding node local allocations to LMB itself, so
it can deal with that during early boot, please use those and stop
trying to re-invent things slightly differently in ways that will not
work with various existing platforms. 

> keep switch smooth. maybe later could try to replace that one by one.
> also i could use lmb_reserve_area() to take name and print out name to
> make debug easy.

I don't want those _area() variants. If you want to pass a name for
debugging purposes, then add it to lmb_reserve() or something like that
and fix the callers.

Cheers,
Ben.

--
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/