diff options
author | Alexey Dobriyan <adobriyan@gmail.com> | 2016-10-07 17:03:12 -0700 |
---|---|---|
committer | Linus Torvalds <torvalds@linux-foundation.org> | 2016-10-07 18:46:30 -0700 |
commit | 81243eacfa400f5f7b89f4c2323d0de9982bb0fb (patch) | |
tree | 9f0bc53a6569fdf1de3d30a61a80386b8bbd1c1f /kernel/groups.c | |
parent | 954f74bf45268bcee0af21b6393c9c8acca7e075 (diff) | |
download | lwn-81243eacfa400f5f7b89f4c2323d0de9982bb0fb.tar.gz lwn-81243eacfa400f5f7b89f4c2323d0de9982bb0fb.zip |
cred: simpler, 1D supplementary groups
Current supplementary groups code can massively overallocate memory and
is implemented in a way so that access to individual gid is done via 2D
array.
If number of gids is <= 32, memory allocation is more or less tolerable
(140/148 bytes). But if it is not, code allocates full page (!)
regardless and, what's even more fun, doesn't reuse small 32-entry
array.
2D array means dependent shifts, loads and LEAs without possibility to
optimize them (gid is never known at compile time).
All of the above is unnecessary. Switch to the usual
trailing-zero-len-array scheme. Memory is allocated with
kmalloc/vmalloc() and only as much as needed. Accesses become simpler
(LEA 8(gi,idx,4) or even without displacement).
Maximum number of gids is 65536 which translates to 256KB+8 bytes. I
think kernel can handle such allocation.
On my usual desktop system with whole 9 (nine) aux groups, struct
group_info shrinks from 148 bytes to 44 bytes, yay!
Nice side effects:
- "gi->gid[i]" is shorter than "GROUP_AT(gi, i)", less typing,
- fix little mess in net/ipv4/ping.c
should have been using GROUP_AT macro but this point becomes moot,
- aux group allocation is persistent and should be accounted as such.
Link: http://lkml.kernel.org/r/20160817201927.GA2096@p183.telecom.by
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Vasily Kulikov <segoon@openwall.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Diffstat (limited to 'kernel/groups.c')
-rw-r--r-- | kernel/groups.c | 67 |
1 files changed, 21 insertions, 46 deletions
diff --git a/kernel/groups.c b/kernel/groups.c index 74d431d25251..2fcadd66a8fd 100644 --- a/kernel/groups.c +++ b/kernel/groups.c @@ -7,55 +7,31 @@ #include <linux/security.h> #include <linux/syscalls.h> #include <linux/user_namespace.h> +#include <linux/vmalloc.h> #include <asm/uaccess.h> struct group_info *groups_alloc(int gidsetsize) { - struct group_info *group_info; - int nblocks; - int i; - - nblocks = (gidsetsize + NGROUPS_PER_BLOCK - 1) / NGROUPS_PER_BLOCK; - /* Make sure we always allocate at least one indirect block pointer */ - nblocks = nblocks ? : 1; - group_info = kmalloc(sizeof(*group_info) + nblocks*sizeof(gid_t *), GFP_USER); - if (!group_info) + struct group_info *gi; + unsigned int len; + + len = sizeof(struct group_info) + sizeof(kgid_t) * gidsetsize; + gi = kmalloc(len, GFP_KERNEL_ACCOUNT|__GFP_NOWARN|__GFP_NORETRY); + if (!gi) + gi = __vmalloc(len, GFP_KERNEL_ACCOUNT|__GFP_HIGHMEM, PAGE_KERNEL); + if (!gi) return NULL; - group_info->ngroups = gidsetsize; - group_info->nblocks = nblocks; - atomic_set(&group_info->usage, 1); - - if (gidsetsize <= NGROUPS_SMALL) - group_info->blocks[0] = group_info->small_block; - else { - for (i = 0; i < nblocks; i++) { - kgid_t *b; - b = (void *)__get_free_page(GFP_USER); - if (!b) - goto out_undo_partial_alloc; - group_info->blocks[i] = b; - } - } - return group_info; -out_undo_partial_alloc: - while (--i >= 0) { - free_page((unsigned long)group_info->blocks[i]); - } - kfree(group_info); - return NULL; + atomic_set(&gi->usage, 1); + gi->ngroups = gidsetsize; + return gi; } EXPORT_SYMBOL(groups_alloc); void groups_free(struct group_info *group_info) { - if (group_info->blocks[0] != group_info->small_block) { - int i; - for (i = 0; i < group_info->nblocks; i++) - free_page((unsigned long)group_info->blocks[i]); - } - kfree(group_info); + kvfree(group_info); } EXPORT_SYMBOL(groups_free); @@ -70,7 +46,7 @@ static int groups_to_user(gid_t __user *grouplist, for (i = 0; i < count; i++) { gid_t gid; - gid = from_kgid_munged(user_ns, GROUP_AT(group_info, i)); + gid = from_kgid_munged(user_ns, group_info->gid[i]); if (put_user(gid, grouplist+i)) return -EFAULT; } @@ -95,7 +71,7 @@ static int groups_from_user(struct group_info *group_info, if (!gid_valid(kgid)) return -EINVAL; - GROUP_AT(group_info, i) = kgid; + group_info->gid[i] = kgid; } return 0; } @@ -115,15 +91,14 @@ static void groups_sort(struct group_info *group_info) for (base = 0; base < max; base++) { int left = base; int right = left + stride; - kgid_t tmp = GROUP_AT(group_info, right); + kgid_t tmp = group_info->gid[right]; - while (left >= 0 && gid_gt(GROUP_AT(group_info, left), tmp)) { - GROUP_AT(group_info, right) = - GROUP_AT(group_info, left); + while (left >= 0 && gid_gt(group_info->gid[left], tmp)) { + group_info->gid[right] = group_info->gid[left]; right = left; left -= stride; } - GROUP_AT(group_info, right) = tmp; + group_info->gid[right] = tmp; } stride /= 3; } @@ -141,9 +116,9 @@ int groups_search(const struct group_info *group_info, kgid_t grp) right = group_info->ngroups; while (left < right) { unsigned int mid = (left+right)/2; - if (gid_gt(grp, GROUP_AT(group_info, mid))) + if (gid_gt(grp, group_info->gid[mid])) left = mid + 1; - else if (gid_lt(grp, GROUP_AT(group_info, mid))) + else if (gid_lt(grp, group_info->gid[mid])) right = mid; else return 1; |