Thread (6 messages) 6 messages, 3 authors, 2014-09-02

Re: [PATCH 1/2] x86: use memblock_alloc_range() or memblock_alloc_base()

From: Akinobu Mita <akinobu.mita@gmail.com>
Date: 2014-08-28 15:50:19
Also in: lkml

2014-08-28 5:53 GMT+09:00 Andrew Morton [off-list ref]:
On Sun, 24 Aug 2014 23:56:02 +0900 Akinobu Mita [off-list ref] wrote:
quoted
Replace memblock_find_in_range() and memblock_reserve() with
memblock_alloc_range() or memblock_alloc_base().
Please spend a little more time preparing the changelogs?
OK, I'll be careful next time.
Why are we making this change?  Because memblock_alloc_range() is
equivalent to memblock_find_in_range()+memblock_reserve() and it's just
a cleanup?  Or is there some deeper functional reason?
This is just a cleanup and I thought there are no functional change.
But I've just realized that the conversion to memblock_alloc_base() in
this patch changes the behaviour in the error case.
Because memblock_alloc_base calls panic if it can't allocate.

So please drop this patch from -mm tree for now.
Does memblock_find_in_range() need to exist?  Can we convert all
callers to memblock_alloc_range()?
There are two callsites where we can't simply convert with
memblock_alloc_range (arch/s390/kernel/setup.c, arch/x86/mm/init.c).

--
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>
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help