Re: [RFC] /sys/kernel/mm/hugepages/hugepages-1048576kB/nr_overcommit_hugepages
From: Michal Hocko <hidden>
Date: 2011-01-05 15:59:54
On Wed 05-01-11 10:36:47, CAI Qian wrote:
From f90b54f9f5607128e375bd78d21e751c433b3cf6 Mon Sep 17 00:00:00 2001 From: CAI Qian <redacted> Date: Wed, 5 Jan 2011 23:26:57 +0800 Subject: [PATCH] hugetlbfs: check invalid nr_hugepages and nr_overcommit_hugepages First, nr_*hugepages* in procfs and sysfs do not check for invalid input like "". Second, when using oversize pages, nr_*hugepages* are expected to be allocated during boot time. Therefore, return -EINVAL for those cases.
I think that the two things should be split into two patches - one for the proper input data handling and the other one for the size check. Albeit, I am not sure about the size check because this is a thing that is just a current implementation limitation and can be change later.
quoted hunk ↗ jump to hunk
Signed-off-by: CAI Qian <redacted> --- fs/sysfs/file.c | 2 ++ mm/hugetlb.c | 18 ++++++++++++++++-- 2 files changed, 18 insertions(+), 2 deletions(-)diff --git a/fs/sysfs/file.c b/fs/sysfs/file.c index da3fefe..9f4ea67 100644 --- a/fs/sysfs/file.c +++ b/fs/sysfs/file.c@@ -207,6 +207,8 @@ flush_write_buffer(struct dentry * dentry, struct sysfs_buffer * buffer, size_t return -ENODEV; rc = ops->store(kobj, attr_sd->s_attr.attr, buffer->page, count); + if (!rc) + return -EINVAL;
This doesn't look correct, you would imbalance sysfs_{get,put}_active.
[...]
--
Michal Hocko
L3 team
SUSE LINUX s.r.o.
Lihovarska 1060/12
190 00 Praha 9
Czech Republic
--
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/ .
Fight unfair telecom policy in Canada: sign http://dissolvethecrtc.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>