Commit Graph

10 Commits

Author SHA1 Message Date
mrg
cb7e28b713 use __diagused where appropriate. 2013-10-30 08:27:01 +00:00
christos
cb2fa8e0d7 Now that KASSERT will always consume its arguments, move a KASSERT inside
a DIAGNOSTIC, because the function used is only defined with DIAGNOSTIC.
2013-10-20 21:12:08 +00:00
hannken
312d89f0de Change bread() and breadn() to never return a buffer on
error and modify all callers to not brelse() on error.

Welcome to 6.99.16

PR kern/46282 (6.0_BETA crash: msdosfs_bmap -> pcbmap -> bread -> bio_doread)
2012-12-20 08:03:41 +00:00
hannken
5d2bff060a Make sure all cached buffers with valid, not yet written data have been
run through copy-on-write.  Call fscow_run() with valid data where possible.

The LP_UFSCOW hack is no longer needed to protect ffs_copyonwrite() against
endless recursion.

- Add a flag B_MODIFY to bread(), breada() and breadn().  If set the caller
  intends to modify the buffer returned.

- Always run copy-on-write on buffers returned from ffs_balloc().

- Add new function ffs_getblk() that gets a buffer, assigns a new blkno,
  may clear the buffer and runs copy-on-write.  Process possible errors
  from getblk() or fscow_run().  Part of PR kern/38664.

Welcome to 4.99.63

Reviewed by: YAMAMOTO Takashi <yamt@netbsd.org>
2008-05-16 09:21:59 +00:00
ad
9f56dfa520 Merge brelse() changes from the vmlocking branch. 2007-10-08 18:02:53 +00:00
rumble
1e80c150c7 Fix a parenthesis bug that would nearly always make the superblock magic
check evaluate to true.
2007-08-06 23:25:27 +00:00
rumble
f486212a0e When computing the superblock checksum, use an unsigned short to avoid an
undesired sign-extension. Now EFS works on big endian machines, too.
2007-07-29 00:56:04 +00:00
rumble
0bbb4e920d Clamp down tigher on the bound around 'numextents'. Add a few comments
to make it clearer what we're tracking.
2007-07-04 20:20:30 +00:00
rumble
0aef936cbe Additional fixes/enhancements:
1) Comply with the way buffercache(9) is intended to be used. Now we
       read in single blocks of EFS_BB_SIZE, never taking in variable
       length extents with a single bread() call.

    2) Handle symlinks with more than one extent. There's no reason for
       this to ever happen, but it's handled now.

    3) Finally, add a hint to our iteration initialiser so we can start
       from the desired offset, rather than naively looping through from
       the beginning each time. Since we can binary search the correct
       location quickly, this improves large sequential reads by about
       40% with 128MB files. Improvement should increase with file size.
2007-07-04 19:24:09 +00:00
rumble
6912898dc8 Add read-only support for SGI's Extent File System.
Reviewed by pooka@.
2007-06-29 23:30:16 +00:00