ipmi: cdev lock_kernel() pushdown

[safe/jmp/linux-2.6] / fs / file.c

diff --git a/fs/file.c b/fs/file.c
index 5110acb..4c6f0ea 100644 (file)
--- a/fs/file.c
+++ b/fs/file.c
@@ -12,6 +12,7 @@
 #include <linux/slab.h>
 #include <linux/vmalloc.h>
 #include <linux/file.h>
+#include <linux/fdtable.h>
 #include <linux/bitops.h>
 #include <linux/interrupt.h>
 #include <linux/spinlock.h>
@@ -149,8 +150,16 @@ static struct fdtable * alloc_fdtable(unsigned int nr)
        nr /= (1024 / sizeof(struct file *));
        nr = roundup_pow_of_two(nr + 1);
        nr *= (1024 / sizeof(struct file *));
-       if (nr > sysctl_nr_open)
-               nr = sysctl_nr_open;
+       /*
+        * Note that this can drive nr *below* what we had passed if sysctl_nr_open
+        * had been set lower between the check in expand_files() and here.  Deal
+        * with that in caller, it's cheaper that way.
+        *
+        * We make sure that nr remains a multiple of BITS_PER_LONG - otherwise
+        * bitmaps handling below becomes unpleasant, to put it mildly...
+        */
+       if (unlikely(nr > sysctl_nr_open))
+               nr = ((sysctl_nr_open - 1) | (BITS_PER_LONG - 1)) + 1;
 
        fdt = kmalloc(sizeof(struct fdtable), GFP_KERNEL);
        if (!fdt)
@@ -199,6 +208,16 @@ static int expand_fdtable(struct files_struct *files, int nr)
        if (!new_fdt)
                return -ENOMEM;
        /*
+        * extremely unlikely race - sysctl_nr_open decreased between the check in
+        * caller and alloc_fdtable().  Cheaper to catch it here...
+        */
+       if (unlikely(new_fdt->max_fds <= nr)) {
+               free_fdarr(new_fdt);
+               free_fdset(new_fdt);
+               kfree(new_fdt);
+               return -EMFILE;
+       }
+       /*
         * Check again since another task may have expanded the fd table while
         * we dropped the lock
         */