CVS commit: src/sys/fs/puffs
Module Name:src
Committed By: pho
Date: Fri Dec 3 13:08:10 UTC 2021
Modified Files:
src/sys/fs/puffs: puffs_msgif.h
Log Message:
Avoid using register_t in
The purpose of this header file is to interface between the
kernel-space and user-space, and is #include'd by a user-space header
. It should therefore not use any of kernel-only types, as
it's not reasonable to require user-land filesystems to #define
_KERNTYPES.
To generate a diff of this commit:
cvs rdiff -u -r1.86 -r1.87 src/sys/fs/puffs/puffs_msgif.h
Please note that diffs are not public domain; they are subject to the
copyright notices on the relevant files.
Modified files:
Index: src/sys/fs/puffs/puffs_msgif.h
diff -u src/sys/fs/puffs/puffs_msgif.h:1.86 src/sys/fs/puffs/puffs_msgif.h:1.87
--- src/sys/fs/puffs/puffs_msgif.h:1.86 Mon Mar 8 17:34:30 2021
+++ src/sys/fs/puffs/puffs_msgif.h Fri Dec 3 13:08:10 2021
@@ -1,4 +1,4 @@
-/* $NetBSD: puffs_msgif.h,v 1.86 2021/03/08 17:34:30 christos Exp $ */
+/* $NetBSD: puffs_msgif.h,v 1.87 2021/12/03 13:08:10 pho Exp $ */
/*
* Copyright (c) 2005, 2006, 2007 Antti Kantee. All Rights Reserved.
@@ -705,7 +705,7 @@ struct puffs_vnmsg_pathconf {
struct puffs_req pvn_pr;
int pvnr_name; /* OUT */
- register_t pvnr_retval; /* IN */
+ __register_t pvnr_retval; /* IN */
};
struct puffs_vnmsg_advlock {
CVS commit: src/sys/fs/puffs
Module Name:src Committed By: pho Date: Fri Dec 3 13:08:10 UTC 2021 Modified Files: src/sys/fs/puffs: puffs_msgif.h Log Message: Avoid using register_t in The purpose of this header file is to interface between the kernel-space and user-space, and is #include'd by a user-space header . It should therefore not use any of kernel-only types, as it's not reasonable to require user-land filesystems to #define _KERNTYPES. To generate a diff of this commit: cvs rdiff -u -r1.86 -r1.87 src/sys/fs/puffs/puffs_msgif.h Please note that diffs are not public domain; they are subject to the copyright notices on the relevant files.
CVS commit: src/sys/fs/puffs
Module Name:src Committed By: christos Date: Fri Sep 27 22:36:57 UTC 2019 Modified Files: src/sys/fs/puffs: puffs_vfsops.c Log Message: Fix copying issue that was causing errors in unit_test puffs_tstavfs by removing code. To generate a diff of this commit: cvs rdiff -u -r1.122 -r1.123 src/sys/fs/puffs/puffs_vfsops.c Please note that diffs are not public domain; they are subject to the copyright notices on the relevant files.
CVS commit: src/sys/fs/puffs
Module Name:src
Committed By: christos
Date: Fri Sep 27 22:36:57 UTC 2019
Modified Files:
src/sys/fs/puffs: puffs_vfsops.c
Log Message:
Fix copying issue that was causing errors in unit_test puffs_tstavfs by
removing code.
To generate a diff of this commit:
cvs rdiff -u -r1.122 -r1.123 src/sys/fs/puffs/puffs_vfsops.c
Please note that diffs are not public domain; they are subject to the
copyright notices on the relevant files.
Modified files:
Index: src/sys/fs/puffs/puffs_vfsops.c
diff -u src/sys/fs/puffs/puffs_vfsops.c:1.122 src/sys/fs/puffs/puffs_vfsops.c:1.123
--- src/sys/fs/puffs/puffs_vfsops.c:1.122 Mon Sep 23 08:00:57 2019
+++ src/sys/fs/puffs/puffs_vfsops.c Fri Sep 27 18:36:57 2019
@@ -1,4 +1,4 @@
-/* $NetBSD: puffs_vfsops.c,v 1.122 2019/09/23 12:00:57 christos Exp $ */
+/* $NetBSD: puffs_vfsops.c,v 1.123 2019/09/27 22:36:57 christos Exp $ */
/*
* Copyright (c) 2005, 2006 Antti Kantee. All Rights Reserved.
@@ -30,7 +30,7 @@
*/
#include
-__KERNEL_RCSID(0, "$NetBSD: puffs_vfsops.c,v 1.122 2019/09/23 12:00:57 christos Exp $");
+__KERNEL_RCSID(0, "$NetBSD: puffs_vfsops.c,v 1.123 2019/09/27 22:36:57 christos Exp $");
#include
#include
@@ -501,14 +501,9 @@ puffs_vfsop_statvfs(struct mount *mp, st
*
* XXX: cache the copy in non-error case
*/
+ copy_statvfs_info(sbp, mp);
if (!error) {
- struct statvfs *sb = STATVFSBUF_GET();
- puffs_statvfs_to_statvfs(_msg->pvfsr_sb, sb);
- copy_statvfs_info(sb, mp);
- STATVFSBUF_PUT(sb);
statvfs_to_puffs_statvfs(sbp, _msg->pvfsr_sb);
- } else {
- copy_statvfs_info(sbp, mp);
}
PUFFS_MSG_RELEASE(statvfs);
Re: CVS commit: src/sys/fs/puffs
On Tue, Oct 07, 2014 at 12:40:04AM +0700, Robert Elz wrote:
Date:Mon, 6 Oct 2014 14:26:44 +
From:Havard Eidnes h...@netbsd.org
Message-ID: 20141006142644.7693...@cvs.netbsd.org
| Make this build again without debugging enabled; DPRINTF() can end up
| as empty, and in an if conditional, you then need braces if that's the
| only potential body.
That change makes no sense to me - the original code was
if (error)
DPRINTF(());
and even if DPRINTF() could produce nothing, the result would be
if (error)
;
which is perfectly good C (useless, but acceptable). Adding { } around
the ';' changes nothing.
Consider what happens if you write:
if (error)
DPRINTF((...));
else
fubar();
When DPRINTF() expands 'if (xxx) yyy' it all goes horribly wrong.
Do we need to support any compilers that don't support __VA_ARGS__ ?
Even microsoft's compiler almost supports it.
David
--
David Laight: da...@l8s.co.uk
Re: CVS commit: src/sys/fs/puffs
On Wed, 15 Oct 2014, David Laight wrote:
Consider what happens if you write:
if (error)
DPRINTF((...));
else
fubar();
When DPRINTF() expands 'if (xxx) yyy' it all goes horribly wrong.
That's why I changed it to
do { if (xxx) yyy; } while (0)
a week or two ago. (That change was not pulled up to netbsd-7,
although other changes made around the same time were pulled up.)
Do we need to support any compilers that don't support __VA_ARGS__ ?
Even microsoft's compiler almost supports it.
I don't know.
--apb (Alan Barrett)
Re: CVS commit: src/sys/fs/puffs
Date:Wed, 15 Oct 2014 19:41:40 +0100
From:David Laight da...@l8s.co.uk
Message-ID: 20141015184140.gl23...@snowdrop.l8s.co.uk
| Consider what happens if you write:
| When DPRINTF() expands 'if (xxx) yyy' it all goes horribly wrong.
Of course, the do { } while(0) wrapping is needed in the DPRINTF
expansion - that was also added - that's there to deal with problems
just like that. WIth that, DPRINTF() acts just like a function
(except it has no value of course) for syntax purposes, and doesn't
need to be surrounded with {} every time it is used.
kre
Re: CVS commit: src/sys/fs/puffs
Date:Mon, 6 Oct 2014 14:26:44 +
From:Havard Eidnes h...@netbsd.org
Message-ID: 20141006142644.7693...@cvs.netbsd.org
| Make this build again without debugging enabled; DPRINTF() can end up
| as empty, and in an if conditional, you then need braces if that's the
| only potential body.
That change makes no sense to me - the original code was
if (error)
DPRINTF(());
and even if DPRINTF() could produce nothing, the result would be
if (error)
;
which is perfectly good C (useless, but acceptable). Adding { } around
the ';' changes nothing.
But in any case, apb changed it yesterday, so DPRINTF() never produces
nothing (personally I'm not sure that should have been needed, but that's
beside the point).
What was the error that resulted in the addition of the {} ?
lre
Re: CVS commit: src/sys/fs/puffs
| Make this build again without debugging enabled; DPRINTF() can end up
| as empty, and in an if conditional, you then need braces if that's the
| only potential body.
That change makes no sense to me - the original code was
if (error)
DPRINTF(());
and even if DPRINTF() could produce nothing, the result would be
if (error)
;
which is perfectly good C (useless, but acceptable). Adding { } around
the ';' changes nothing.
Well...
compile puffs/puffs_vnops.o
/u/build/HEAD/src/sys/fs/puffs/puffs_vnops.c: In function 'zerofill_lastpage':
/u/build/HEAD/src/sys/fs/puffs/puffs_vnops.c:1160:70: error: suggest braces
around empty body in an 'if' statement [-Werror=empty-body]
DPRINTF((zero-fill 0x%lx@0x% PRIx64 = %d\n, len, off, error));
^
cc1: all warnings being treated as errors
*** Failed target: puffs_vnops.o
But in any case, apb changed it yesterday, so DPRINTF() never produces
nothing (personally I'm not sure that should have been needed, but that's
beside the point).
Ah. It's quite possible that my latest update was from before
that commit.
What was the error that resulted in the addition of the {} ?
See above.
Regards,
- Håvard
Re: CVS commit: src/sys/fs/puffs
On Sun, Oct 05, 2014 at 02:13:15PM +, Alan Barrett wrote:
#ifdef PUFFSDEBUG
extern int puffsdebug; /* puffs_subr.c */
-#define DPRINTF(x) if (puffsdebug 0) printf x
-#define DPRINTF_VERBOSE(x) if (puffsdebug 1) printf x
+#define DPRINTF(x) do { \
+ if (puffsdebug 0) printf x; \
+ while (/*CONSTCOND*/0)
I think it'd be even more safe to close the block with a '}' before the
while.
+#define DPRINTF_VERBOSE(x) do { \
+ if (puffsdebug 1) printf x; \
+ while (/*CONSTCOND*/0)
Here too.
#else
-#define DPRINTF(x)
-#define DPRINTF_VERBOSE(x)
+#define DPRINTF(x) ((void)0)
+#define DPRINTF_VERBOSE(x) ((void)0)
#endif
#define MPTOPUFFSMP(mp) ((struct puffs_mount *)((mp)-mnt_data))
(I must admit, not compile-tested)
Regards,
Alistair
Re: CVS commit: src/sys/fs/puffs
David Holland dholland-sourcechan...@netbsd.org wrote: Di not call cache_enter with path components bigger than NCHNAMLEN, as it panics the kernel. This is wrong; you're reverting rmind's cleanup. I backed out the change. Will someone pullup the change to netbsd-6? -- Emmanuel Dreyfus http://hcpnet.free.fr/pubz m...@netbsd.org
Re: CVS commit: src/sys/fs/puffs
On Mon, Jul 23, 2012 at 10:07:06PM +, Martin Husemann wrote: This is wrong; you're reverting rmind's cleanup. I backed out the change. Will someone pullup the change to netbsd-6? Can you prepare a patch that applies before rmind's cleanup? (Or just one that applies on netbsd-6?) OTOH for netbsd-6 maybe a better patch would be to change cache_enter to return instead of asserting for long filenames. (not sure if this is related) Unfortunately even after the backout a test run still fails reproducably for me - see kern/46734. Doesn't look particularly related to me. Do you know which test kills it? (I guess post it to the PR, not here...) -- David A. Holland dholl...@netbsd.org
Re: CVS commit: src/sys/fs/puffs
David Holland dholland-sourcechan...@netbsd.org wrote: OTOH for netbsd-6 maybe a better patch would be to change cache_enter to return instead of asserting for long filenames. If nobody opposes, I can do that. -- Emmanuel Dreyfus http://hcpnet.free.fr/pubz m...@netbsd.org
Re: CVS commit: src/sys/fs/puffs
On Tue, Jul 24, 2012 at 07:13:34AM +0200, Emmanuel Dreyfus wrote: OTOH for netbsd-6 maybe a better patch would be to change cache_enter to return instead of asserting for long filenames. If nobody opposes, I can do that. Best to ask on tech-kern, but it seems unlikely to be controversial. It's hiding bugs but only in the sense that the interface was stupid; and it's probable that puffs isn't the only fs missing the check. -- David A. Holland dholl...@netbsd.org
Re: CVS commit: src/sys/fs/puffs
On Tue, Oct 18, 2011 at 03:39:10PM +, Emmanuel Dreyfus wrote: Module Name: src Committed By: manu Date: Tue Oct 18 15:39:09 UTC 2011 Modified Files: src/sys/fs/puffs: puffs_msgif.c puffs_node.c puffs_vfsops.c puffs_vnops.c Log Message: Make sure pagedaemon does not sleep for memory in puffs_vnop_sleep. Add KASSERT on any sleeping memory allocation to check it cannot happen again. +#ifdef DIAGNOSTIC + KASSERT(curlwp != uvm.pagedaemon_lwp); +#endif Why pollute code with redundant #ifdefs? - Jukka.
Re: CVS commit: src/sys/fs/puffs
Jukka Ruohonen jruoho...@iki.fi wrote: +#ifdef DIAGNOSTIC + KASSERT(curlwp != uvm.pagedaemon_lwp); +#endif Why pollute code with redundant #ifdefs? Right, I missed the point that KASSERT contains #ifdef DIAGNOSTIC. I will fix that tomorrow morning. -- Emmanuel Dreyfus http://hcpnet.free.fr/pubz m...@netbsd.org
Re: CVS commit: src/sys/fs/puffs
hi, YAMAMOTO Takashi y...@mwd.biglobe.ne.jp wrote: if it trapped forever, it's a bug and should be fixed. my point was that your change didn't fix the bug. blocking ioflush is merely a symptom. The problem with userland filesystems is that we may have little control as theses may be third pary programs. Should kernel threads trust theses processes in order to run as intended? i guess it depends. is ioflush blocking forever worse than a userland application blocking forever? it depends. even if it's a problem, it isn't a problem specific to userland filesystems. filesystems have varying performance and reliability. One way to fix that may be to have one ioflush thread for each userland filesystem. That way a broken filesystem will not prevent ioflush from working for others. But we have hit similar problems with others kernel threads. sure, it can be a good idea. YAMAMOTO Takashi -- Emmanuel Dreyfus http://hcpnet.free.fr/pubz m...@netbsd.org
Re: CVS commit: src/sys/fs/puffs
On Wed, Oct 12, 2011 at 06:02:05AM +, YAMAMOTO Takashi wrote: One way to fix that may be to have one ioflush thread for each userland filesystem. That way a broken filesystem will not prevent ioflush from working for others. sure, it can be a good idea. I have a patch that does that, I will clean it up and post it. -- Emmanuel Dreyfus m...@netbsd.org
Re: CVS commit: src/sys/fs/puffs
hi, YAMAMOTO Takashi y...@mwd.biglobe.ne.jp wrote: Sure ioflush do not directly free memory, but vnodes' dirty page use memory, don't they? If ioflush stops working, is pageadaemon able to pageout that kind of memory? pagedaemon flushes dirty pages by itself, yes. So this is not a problem to get ioflush traped forever awaiting for a VOP_FSYNC completion? There must be some drawback, otherwise ioflush would be useless. if it trapped forever, it's a bug and should be fixed. my point was that your change didn't fix the bug. blocking ioflush is merely a symptom. on the other hand, it needlessly introduced a limitation on what kernel threads can do. now we have vdrain_thread which i think can hit the LW_SYSTEM assertion in puffs_vnop_fsync. a possible solution would be local page recycling. ie. reserve some pages and put them onto a page queue dedicated for a given set of processes so that pages can be recycled in the set independently from the global queue. But we do not know how much memory the userland fileserver is going to require. We can have some reserves, but we must be ready to kill pigs if too much memory is retained by the fileserver. HEAD does it, but netbsd-5 prefers to hang. you don't need to know how much memory the server will use. when the server runs out of the reserved set of pages, you can reuse pages in the set. YAMAMOTO Takashi -- Emmanuel Dreyfus http://hcpnet.free.fr/pubz m...@netbsd.org
Re: CVS commit: src/sys/fs/puffs
YAMAMOTO Takashi y...@mwd.biglobe.ne.jp wrote: if it trapped forever, it's a bug and should be fixed. my point was that your change didn't fix the bug. blocking ioflush is merely a symptom. The problem with userland filesystems is that we may have little control as theses may be third pary programs. Should kernel threads trust theses processes in order to run as intended? One way to fix that may be to have one ioflush thread for each userland filesystem. That way a broken filesystem will not prevent ioflush from working for others. But we have hit similar problems with others kernel threads. -- Emmanuel Dreyfus http://hcpnet.free.fr/pubz m...@netbsd.org
Re: CVS commit: src/sys/fs/puffs
On Sat, Sep 24, 2011 at 03:31:02AM +, YAMAMOTO Takashi wrote: pagedaemon flushes dirty pages by itself, yes. I added a printf() at the beginning of puffs_vnop_strategy() to check for calls by kernel threads. Only ioflush calls puffs_vnop_strategy(), pagedaemon never goes there. This may explain why the system cannot cope with PUFFS and memory shortage. An interesting point: if I use PDPOLICY_CLOCKPRO, then pagedaemon does call puffs_vnop_strategy(), through the path below. I understand this is the way it is supposed to work: puffs_vnop_strategy VOP_STRATEGY genfs_do_io genfs_gop_write genfs_do_putpages genfs_putpages VOP_PUTPAGES (through pgo-pgo_put) uvm_pageout The system still gets stuck after a while, with glusterfsd sleeping forever in flt_noram1. ddb' show uvmexp shows 3 free pages, 509 paging. If I understand correctly, pagedaemon performs an asynchronous VOP_STRATEGY and count an operation stuck in the PUFFS filesystem as paging. It decides to do more work if uvmexp.free + uvmexp.paging gets too low. therefore with many PUFFS page forever in the paging count, pagedaemon will not help us. I have a patch that maintain a count of paging to remote filesystems (that is, the one that do not have MNT_LOCAL in struct mount mnt_flag), and substract it from uvmexp.paging when pagedaemon has to decide whether to send more pages to the swap. It seems to work fine, but the system hang with perfused awaiting for km_getwait2 (from sys_write/putter_fop_write) -- Emmanuel Dreyfus m...@netbsd.org
Re: CVS commit: src/sys/fs/puffs
YAMAMOTO Takashi y...@mwd.biglobe.ne.jp wrote: This avoids deadlocks in the following situations: 1) when memory is low: ioflush waits the fileystem, the fielsystem waits for memory can you explain how it is a problem? As I understand, one way to free memory is to flush vnode backed pages to the backend storage. If ioflush calls VOP_FSYNC on a memory-starved userland filesystem, it will get stuck until the filesystem gets memory again, and while it is stuck, it does not help freeing memory. -- Emmanuel Dreyfus http://hcpnet.free.fr/pubz m...@netbsd.org
Re: CVS commit: src/sys/fs/puffs
hi, YAMAMOTO Takashi y...@mwd.biglobe.ne.jp wrote: This avoids deadlocks in the following situations: 1) when memory is low: ioflush waits the fileystem, the fielsystem waits for memory can you explain how it is a problem? As I understand, one way to free memory is to flush vnode backed pages to the backend storage. If ioflush calls VOP_FSYNC on a memory-starved userland filesystem, it will get stuck until the filesystem gets memory again, and while it is stuck, it does not help freeing memory. as i told you a few times, ioflush is not a thread to free memory. pagedaemon is. please read ufs_bmaparray and grep uvm.pagedaemon_lwp in src/sys/kern/ to see what other filesystems do. i don't think this approach works for puffs because it's almost impossible to say in which cases an operation needs memory allocation to complete, though. YAMAMOTO Takashi -- Emmanuel Dreyfus http://hcpnet.free.fr/pubz m...@netbsd.org
Re: CVS commit: src/sys/fs/puffs
YAMAMOTO Takashi y...@mwd.biglobe.ne.jp wrote: as i told you a few times, ioflush is not a thread to free memory. pagedaemon is. Sure ioflush do not directly free memory, but vnodes' dirty page use memory, don't they? If ioflush stops working, is pageadaemon able to pageout that kind of memory? please read ufs_bmaparray and grep uvm.pagedaemon_lwp in src/sys/kern/ to see what other filesystems do. i don't think this approach works for puffs because it's almost impossible to say in which cases an operation needs memory allocation to complete, though. It is indeed impossible, as we cannot know what the userland filesystem will do. -- Emmanuel Dreyfus http://hcpnet.free.fr/pubz m...@netbsd.org
Re: CVS commit: src/sys/fs/puffs
hi, YAMAMOTO Takashi y...@mwd.biglobe.ne.jp wrote: as i told you a few times, ioflush is not a thread to free memory. pagedaemon is. Sure ioflush do not directly free memory, but vnodes' dirty page use memory, don't they? If ioflush stops working, is pageadaemon able to pageout that kind of memory? pagedaemon flushes dirty pages by itself, yes. please read ufs_bmaparray and grep uvm.pagedaemon_lwp in src/sys/kern/ to see what other filesystems do. i don't think this approach works for puffs because it's almost impossible to say in which cases an operation needs memory allocation to complete, though. It is indeed impossible, as we cannot know what the userland filesystem will do. a possible solution would be local page recycling. ie. reserve some pages and put them onto a page queue dedicated for a given set of processes so that pages can be recycled in the set independently from the global queue. YAMAMOTO Takashi -- Emmanuel Dreyfus http://hcpnet.free.fr/pubz m...@netbsd.org
Re: CVS commit: src/sys/fs/puffs
YAMAMOTO Takashi y...@mwd.biglobe.ne.jp wrote: Sure ioflush do not directly free memory, but vnodes' dirty page use memory, don't they? If ioflush stops working, is pageadaemon able to pageout that kind of memory? pagedaemon flushes dirty pages by itself, yes. So this is not a problem to get ioflush traped forever awaiting for a VOP_FSYNC completion? There must be some drawback, otherwise ioflush would be useless. a possible solution would be local page recycling. ie. reserve some pages and put them onto a page queue dedicated for a given set of processes so that pages can be recycled in the set independently from the global queue. But we do not know how much memory the userland fileserver is going to require. We can have some reserves, but we must be ready to kill pigs if too much memory is retained by the fileserver. HEAD does it, but netbsd-5 prefers to hang. -- Emmanuel Dreyfus http://hcpnet.free.fr/pubz m...@netbsd.org
Re: CVS commit: src/sys/fs/puffs
hi, Module Name: src Committed By: manu Date: Wed Sep 21 15:36:33 UTC 2011 Modified Files: src/sys/fs/puffs: puffs_vfsops.c puffs_vnops.c Log Message: Make sure ioflush does not sleep in PUFFS code path, waiting for a mutex, a memory allocation, or a response from the filesystem. This avoids deadlocks in the following situations: 1) when memory is low: ioflush waits the fileystem, the fielsystem waits for memory can you explain how it is a problem? YAMAMOTO Takashi 2) when the filesystem does not respond (e.g.: network outage ona distributed filesystem) To generate a diff of this commit: cvs rdiff -u -r1.96 -r1.97 src/sys/fs/puffs/puffs_vfsops.c cvs rdiff -u -r1.155 -r1.156 src/sys/fs/puffs/puffs_vnops.c Please note that diffs are not public domain; they are subject to the copyright notices on the relevant files.
Re: CVS commit: src/sys/fs/puffs
On Tue Jan 11 2011 at 14:04:54 +, Mihai Chelaru wrote: Module Name: src Committed By: kefren Date: Tue Jan 11 14:04:54 UTC 2011 Modified Files: src/sys/fs/puffs: puffs_sys.h puffs_vnops.c Log Message: add advlock to puffs. ok pooka@ should fix kern/43321 Thanks! This is a big improvement and fixes the problem in most cases. However, e.g. for distributed file systems it would be good to have an additional mode where the server can decide the locking policy. I'll bump the PR state to analyzed to indicate progress. -- älä karot toivorikkauttas, kyl rätei ja lumpui piisaa
Re: CVS commit: src/sys/fs/puffs
On Wed, Jul 14, 2010 at 02:07:37PM +, Antti Kantee wrote: Modified Files: src/sys/fs/puffs: puffs_vnops.c Log Message: RENAME lookup semantics say return EISDIR if dvp = *vpp for the last component obviously(!!) I take it that when you say *vpp, you mean tvp? (the alphabet soup effect is why I have been looking for better names for this crud) also I think the case you actually changed should be EINVAL? -- David A. Holland dholl...@netbsd.org
Re: CVS commit: src/sys/fs/puffs
On Thu Jul 15 2010 at 20:03:37 +, David Holland wrote: On Wed, Jul 14, 2010 at 02:07:37PM +, Antti Kantee wrote: Modified Files: src/sys/fs/puffs: puffs_vnops.c Log Message: RENAME lookup semantics say return EISDIR if dvp = *vpp for the last component obviously(!!) I take it that when you say *vpp, you mean tvp? Of course not. That error comes from lookup to avoid do_sys_rename() overriding it. (the alphabet soup effect is why I have been looking for better names for this crud) also I think the case you actually changed should be EINVAL? I don't care as long as they're all the same (and nothing breaks ;).
CVS commit: src/sys/fs/puffs
Module Name:src Committed By: pooka Date: Thu Jan 14 14:44:13 UTC 2010 Modified Files: src/sys/fs/puffs: puffs_vnops.c Log Message: Since VOP_GETATTR() does not require a locked vnode, resolve and reference the puffs_node before sending the request to the file server. This diminishes the window where the inode can be reclaimed and be invalidated before it is accessed (but does not completely eliminate the race, as that is a caller problem which we cannot fix here). To generate a diff of this commit: cvs rdiff -u -r1.141 -r1.142 src/sys/fs/puffs/puffs_vnops.c Please note that diffs are not public domain; they are subject to the copyright notices on the relevant files.
CVS commit: src/sys/fs/puffs
Module Name:src Committed By: pooka Date: Thu Jan 7 22:45:31 UTC 2010 Modified Files: src/sys/fs/puffs: puffs_msgif.c puffs_msgif.h puffs_sys.h Log Message: Add a PUFFS_UNMOUNT server-kernel request, which causes the kernel to initiate self destruct, i.e. unmount(MNT_FORCE). This, however, is a semi-controlled self-destruct, since all caches are flushed before the (possibly) violent unmount takes place. To generate a diff of this commit: cvs rdiff -u -r1.76 -r1.77 src/sys/fs/puffs/puffs_msgif.c cvs rdiff -u -r1.68 -r1.69 src/sys/fs/puffs/puffs_msgif.h cvs rdiff -u -r1.73 -r1.74 src/sys/fs/puffs/puffs_sys.h Please note that diffs are not public domain; they are subject to the copyright notices on the relevant files.
CVS commit: src/sys/fs/puffs
Module Name:src Committed By: pooka Date: Thu Jan 7 22:59:27 UTC 2010 Modified Files: src/sys/fs/puffs: puffs_msgif.c Log Message: Fix variable name in my commit tree too. To generate a diff of this commit: cvs rdiff -u -r1.77 -r1.78 src/sys/fs/puffs/puffs_msgif.c Please note that diffs are not public domain; they are subject to the copyright notices on the relevant files.
