t; cvs rdiff -u -r1.62 -r1.63 src/sys/fs/cd9660/cd9660_vnops.c
First, this is not backwards compatible. An old mount_cd9660 and
a new kernel will always return EINVAL as the argument length
check on top of cd9660_mount() fires.
Second, with these new mount arguments an update mount has to update
t
> On 3. Feb 2024, at 18:38, Christos Zoulas wrote:
>
> Fixed, thanks.
Not really, with "struct iso_args *args = data" you cannot set fields
beyond "datalen" here and the MNT_UPDATE case is still missing.
--
J. Hannken-Illjes - hann...@mailbox.org
.. 312
of cd9660_vfsops.c should test them and fail on different values.
If it is allowed to change them it has to be done here.
>> On Feb 3, 2024, at 12:48 PM, J. Hannken-Illjes wrote:
>>
>>> On 3. Feb 2024, at 18:38, Christos Zoulas wrote:
>>>
>>> Fixed, t
gt; cvs rdiff -u -r1.406 -r1.407 src/sys/dev/raidframe/rf_netbsdkintf.c
>
> Please note that diffs are not public domain; they are subject to the
> copyright notices on the relevant files.
Thanks, replacing vput() with vrele() would have been even better ...
--
J. Hannken-Illjes - hann.
ith say 20 knobs,
do we build all 2**20 configurations to be sure everything at
least builds?
Isn't it better to always build everything and move this selection
into the set lists or whatever you use to get the final image?
--
J. Hannken-Illjes - hann...@mailbox.org
signature.asc
Description: Message signed with OpenPGP
o the /dev/xxx nodes.
This breaks at least sparc64/GENERIC:
sys/dev/sequencer.c:285:1: error: no previous prototype for 'sequencerattach'
[-Werror=missing-prototypes]
285 | sequencerattach(int n)
| ^~~~~~~~~~~
It has "midi* at midibus?" but no "pseudo-devic
s from /etc/fstab
> first, and if /sbin/mount fails, then try to find a zfs dataset
> for the failed entry and /sbin/zfs mount it. But if things go
> wrong, a complicated mounting process will make troubleshooting
> harder. For that reason, I'd like to keep mountcrit_zfs separate
> from mountcrit_local.
--
J. Hannken-Illjes - hann...@mailbox.org
signature.asc
Description: Message signed with OpenPGP
ngular LWP.
>
>
> To generate a diff of this commit:
...
> cvs rdiff -u -r1.63 -r1.64 src/sys/kern/sys_select.c
-sleepq_unsleep(l, false);
+sleepq_remove(l->l_sleepq, l, true);
}
}
mutex_spin_exit(lock);
Looks like sleepq_remove() unlocks l->l_mutex == lock and
t
> On 10. Oct 2023, at 20:58, Andrew Doran wrote:
>
> On Tue, Oct 10, 2023 at 06:00:57PM +0200, J. Hannken-Illjes wrote:
>
>>> cvs rdiff -u -r1.63 -r1.64 src/sys/kern/sys_select.c
>>
>> -sleepq_unsleep(l, false);
>>
ops.c
This is completely wrong!
The call sequence is "VOP_PATHCONF -> zfs_netbsd_pathconf -> zfs_pathconf"
where zfs_netbsd_pathconf() handles the NetBSD specific names if
zfs_pathconf() returns EOPNOTSUPP and also returns EINVAL for
unsupported names in this case.
Please revert.
--
J. Han
Frank,
the vnode operations for ".zfs" are in zfs_ctldir.c and there
is no pathconf operation here ...
--
J. Hannken-Illjes - hann...@mailbox.org
> On 3. Oct 2022, at 11:26, Frank Kardel wrote:
>
> Well, I am am happy to do that. But ls got eopnotsupp for fpathconf of ACL
Yes -- I suppose genfs_pathconf() is sufficient here.
--
J. Hannken-Illjes - hann...@mailbox.org
> On 3. Oct 2022, at 11:40, Frank Kardel wrote:
>
>
> Good to know. So do we need to add path conf there?
> Frank
>
> 3 Oct 2022 11:34:09 J. Hannken-Illjes :
>
&g
Looks like an error in the `list' file:
SPECIAL pingsrcdir distrib/utils/x_gzip
SPECIAL pingsrcdir distrib/utils/x_ping
--
Juergen Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
On Jun 21, 2011, at 1:36 AM, Christos Zoulas wrote:
On Aug 29, 2011, at 2:39 AM, Mindaugas Rasiukevicius wrote:
> Module Name: src
> Committed By: rmind
> Date: Mon Aug 29 00:39:16 UTC 2011
>
> Modified Files:
> src/sys/kern: sys_select.c
>
> Log Message:
> Add kern.direct_select sysctl. Default to 0 for now.
Why is it disabled
On Sep 29, 2011, at 12:52 AM, Matt Thomas wrote:
> Module Name: src
> Committed By: matt
> Date: Wed Sep 28 22:52:16 UTC 2011
>
> Modified Files:
> src/sys/uvm: uvm_page.c uvm_pager.c uvm_pager.h
>
> Log Message:
> Reallocate emergency pager va when ncolors is increased. (modicat
On Oct 2, 2011, at 3:26 PM, Adam Hamsik wrote:
>
> On Oct,Sunday 2 2011, at 3:00 PM, Juergen Hannken-Illjes wrote:
>
>> Module Name: src
>> Committed By:hannken
>> Date:Sun Oct 2 13:00:07 UTC 2011
>>
>> Modified Files:
>> src/sys/kern: vfs_vnode.c
>>
>> Log Messag
On Oct 8, 2011, at 12:42 AM, Roland Dowdeswell wrote:
> Module Name: src
> Committed By: elric
> Date: Fri Oct 7 22:42:19 UTC 2011
>
> Modified Files:
> src/sys/dev/pci: if_iwn.c
This breaks building kernels:
sys/dev/pci/if_iwn.c: In function 'iwn5000_runtime_calib':
sys/dev/pci
On Oct 15, 2011, at 3:30 AM, matthew green wrote:
>
>> Module Name: src
>> Committed By:hannken
>> Date:Fri Oct 14 09:23:31 UTC 2011
>>
>> Modified Files:
>> src/sys/compat/linux/common: linux_file.c linux_file64.c linux_ioctl.c
>> linux_misc.c
>> src/
On Nov 15, 2011, at 4:08 AM, YAMAMOTO Takashi wrote:
> hi,
>
>> Module Name: src
>> Committed By:hannken
>> Date:Sun Oct 30 12:00:28 UTC 2011
>>
>> Modified Files:
>> src/sys/nfs: nfs_serv.c
>>
>> Log Message:
>> VOP_GETATTR() needs a shared lock at least.
>
> this
On Oct 31, 2011, at 1:42 PM, YAMAMOTO Takashi wrote:
> Module Name: src
> Committed By: yamt
> Date: Mon Oct 31 12:42:53 UTC 2011
>
> Modified Files:
> src/sys/arch/i386/i386: dumpsys.c
>
> Log Message:
> dumpsys_seg: don't overwrite the previous mapping
With this change in plac
On Nov 19, 2011, at 5:11 PM, Christos Zoulas wrote:
> Module Name: src
> Committed By: christos
> Date: Sat Nov 19 16:11:24 UTC 2011
>
> Modified Files:
> src/sys/kern: cnmagic.c
>
> Log Message:
> PR/45633: Christian Biere: Don't access byte after NUL when setting magic.
This c
On Nov 20, 2011, at 6:10 PM, Christos Zoulas wrote:
> On Nov 20, 10:02am, hann...@eis.cs.tu-bs.de ("J. Hannken-Illjes") wrote:
> -- Subject: Re: CVS commit: src/sys/kern
>
> |
> | On Nov 19, 2011, at 5:11 PM, Christos Zoulas wrote:
> |
> | > Module
On Nov 23, 2011, at 11:11 AM, Alan Barrett wrote:
>>> My vote would be to remove [unionfs]; it doesn't work and the only reason
>>> it was ever brought in had to do with alleged locking improvements.
>>
>> Is anyone using it?
>
> I used to make heavy use of unionfs, and I had no problems. (Th
On Feb 21, 2012, at 2:47 AM, Jonathan A. Kollasch wrote:
> Module Name: src
> Committed By: jakllsch
> Date: Tue Feb 21 01:47:50 UTC 2012
>
> Modified Files:
> src/sys/arch/xen/xen: xbd_xenbus.c
>
> Log Message:
> Add and use xbdminphys() to handle transfer segmentation/size limit
On Nov 3, 2012, at 3:44 PM, Izumi Tsutsui wrote:
> Module Name: src
> Committed By: tsutsui
> Date: Sat Nov 3 14:44:29 UTC 2012
>
> Modified Files:
> src/sys/arch/x68k/conf: GENERIC
>
> Log Message:
> Comment out pseudo-device fss. We already has options FFS_NO_SNAPSHOT.
This i
On Nov 3, 2012, at 4:51 PM, Izumi Tsutsui wrote:
> hannken@ wrote:
>
>>> Modified Files:
>>> src/sys/arch/x68k/conf: GENERIC
>>>
>>> Log Message:
>>> Comment out pseudo-device fss. We already has options FFS_NO_SNAPSHOT.
>>
>> This is not the same:
>>
>> - having fss makes it possible to
they supposed to leave it in an undefined state? Either way, it seems
> to me that the intent ought to be documented in buffercache(9).
Both bread() and breadn() will either return with `error == 0, *bpp != NULL'
or `error != 0, *bpp == NULL'.
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
; but there is no I/O pending.
> This caused my laptop to hang on boot_findwedge because:
>findroot: unable to read block 358331527 of dev dk0 (22)
Thanks. Just applied the same to breadn().
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
unmount() and do_sys_sync() for example.
The lock order is mountlist_lock->mnt_unmounting, but dounmount()
locks in reverse direction.
Please fix or revert.
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
On Apr 28, 2013, at 10:56 AM, Michael van Elst wrote:
> On Sun, Apr 28, 2013 at 10:31:19AM +0200, J. Hannken-Illjes wrote:
>> On Apr 27, 2013, at 12:27 AM, Michael van Elst wrote:
>>
>>> Module Name:src
>>> Committed By: mlelstv
>>> D
On Apr 28, 2013, at 4:32 PM, Michael van Elst wrote:
> On Sun, Apr 28, 2013 at 01:50:37PM +0200, J. Hannken-Illjes wrote:
>
>>> The locking order hasn't been changed, but the faulty tryenter has been
>>> replaced with an enter.
>>
>> Sure -- and th
PR/41068.
> Thanks to wiz@ for stress testing.
The tests fs/vfs/t_union/tmpfs_basic and fs/vfs/t_union/tmpfs_whiteout
start failing after this commit.
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
n and protects "vclean(); vrelel()" and "vrelel()"
>> against "vget()".
>>
>> Presented on tech-kern.
>
> There are multiple reports of hangs in VFS (see "something's wrong" thread
> on current-users@ and PR/48411). It seems
ed
/etc/ntp.conf fails as:
ntpd[317]: line 33 column 1 syntax error, unexpected T_String, expecting $end
ntpd[317]: syntax error in /etc/ntp.conf line 33, column 1
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
in more detail. Tmpfs_fhtovp() will fail
as soon as an unlinked tmpfs node drops its last vnode reference.
> Why not just check for tn_links == 0 in tmpfs_fhtovp()?
Because it is ok as long as the corresponding vnode is open/referenced.
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
On Jan 3, 2014, at 10:18 PM, J. Hannken-Illjes wrote:
> On Jan 3, 2014, at 10:13 PM, Mindaugas Rasiukevicius wrote:
>
>> "Juergen Hannken-Illjes" wrote:
>>> Module Name:src
>>> Committed By: hannken
>>> Date: Fri
ents(), thus saving
> calls to kmem_zalloc() and kmem_free(); OK rmind@. From OpenBSD.
Is it really a good idea to allocate 528 bytes on the kernel stack?
File systems nest and already use much stack space.
Looks better to use a pool_cache.
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
quad support.
This change breaks sparc64:
nbmake[8]:
"src/sys/arch/sparc/stand/ofwboot/../../../../lib/libkern/../../../common/lib/libc/Makefile.inc"
line 6: Malformed conditional (${HAVE_LIBGCC} != "no")
src/sys/arch/sparc/stand/ofwboot/Makefile includes
..../common/lib/
KGCCCMDS must be no).
This breaks all builds as MKGCCCMDS is now missing from the set list
print controls.
See distrib/sets/mkvars.mk, target mkvars
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
lic domain; they are subject to the
> copyright notices on the relevant files.
Great! Please request pullups to -8 and -9.
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig
signature.asc
Description: Message signed with OpenPGP
by the
calling thread, shared lock held by anyone or unlocked,
respectively.
This function must never be used to make locking decisions at
run time: it is provided only for diagnostic purposes.
I suppose you cannot determine if the current thread holds
a sha
> On Feb 17, 2020, at 11:19 PM, Andrew Doran wrote:
>
> Hi,
>
> On Thu, Feb 06, 2020 at 06:33:55PM +0100, J. Hannken-Illjes wrote:
>
>>> On 12. Jan 2020, at 18:49, Andrew Doran wrote:
>>>
>>> Module Name:src
>>> Committed By:
/zfs/metaslab.c
>>
>> Please note that diffs are not public domain; they are subject to the
>> copyright notices on the relevant files.
>>
>
> Forgot to add in the commit log, the changes have been pulled in from
> upstream openzfs.
>
> https://github.com
rump/rumpkern/rump.c:426
Are sure it is ok to detach a non-attached event?
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig
signature.asc
Description: Message signed with OpenPGP
esn't reference the
vnode/znode and works for in-memory znodes only so we have to guarantee
it doesn't get called after VOP_RECLAIM().
> Do you have a test case that can trigger the problem?
Under load I get use-after-free of znodes in zfs_get_data() or entirely
miss znodes her
the relevant files.
This broke most or all NFS tests on ATF (see attached list).
Please revert or investigate.
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
fs/nfs/t_mountd::mountdhup
fs/vfs/t_full::nfs_fillfs
fs/vfs/t_io::nfs_extendfile
fs/vfs/t_io::nfs_extendfile_app
apable: on
Per-Vector Masking Capable: off
Extended Message Data Capable: off
Extended Message Data Enable: off
Message Address (lower) register: 0xfee0
Message Address (upper) register: 0x
Message Data register: 0x0064
Any ideas how to fix this issue?
-
Jared,
please test if the attached diff is sufficient.
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig
uvm_swap.c.diff
Description: Binary data
> On 17. Feb 2021, at 13:07, Jared McNeill wrote:
>
> I noticed this on reboot since this change:
>
> [ 3636.2
ant sleep and operations going down
to specfs for example. Here we have the problem that the specfs
operation may take forever -- reading from a tty for example.
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
ile is obsolete or flist is out of date ?
--
./usr/include/g++/bits/sparc
./usr/include/g++/bits/sparc64
= end of 2 extra files ===
Reverting this change makes sparc64 build again.
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
27;t protect it ...
As this commit prevents other threads from obtaining a lock while
VOP_RECLAIM() is running the protection should be the same as before.
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
On Feb 27, 2014, at 7:01 PM, Mindaugas Rasiukevicius wrote:
> "J. Hannken-Illjes" wrote:
>>> I have not had time to follow your VFS changes, but can you explain why
>>> did you remove VOP_LOCK/VOP_UNLOCK in tmpfs_reclaim()? It was added to
>>>
DESTDIR.
File wasn't installed ?
--
./usr/include/g++/bits/sparc/c++config.h
./usr/include/g++/bits/sparc64/c++config.h
==== end of 2 missing files ==
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
ck tick the softclk thread calls timer_intr() which
wants the proc_lock. Now we successfully stopped all softclk threads.
The callout for kpause() will not fire -> deadlock/
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
. Use consistent locking around the per-cpu data.
This change leads to a deadlock:
Thread 1 calls cache_revlookup(),
holds ncp->nc_lock @ line 589,
wants cpup->cpu_lock @ line 604
Thread 2 calls cache_lookup(),
holds cpup->cpu_lock @ line 391,
calls cache_lookup_entry(),
w
This cannot work.
We now run cache_lookup_entry() without a cpu_lock and without
namecache_lock so the list may change while it gets traversed.
See the comment on top of cache_lookup_entry().
Please revert your changes to vfs_cache.c.
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
or: a parameter list without
> types is only allowed in a function definition
> VFS_PROTOS(msdosfs);
Maybe a patch to lsof/dialects/n+obsd/dlsof.h like
#if defined (NETBSDV)
#define VFS_PROTOS(dummy) /**/
#endif
before including msdosfsmount.h?
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
nttypes.c:173:2: error: format '%hhd' expects argument of
type 'signed char *', but argument 3 has type 'int_fast8_t *' [-Werror=format=]
Kannst Du Dir das bitte einmal ansehen?
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
On 25 Jul 201
oops, this was meant for Joerg only ...
Since this commit i386 doesn't build, pleas take a look.
On 28 Jul 2014, at 09:40, J. Hannken-Illjes wrote:
> Jörg,
>
> seit diesem Commit gibt es keine i386 Builds mehr:
>
> src/tests/include/t_inttypes.c:173:2: error: format &#x
ead/pthread_tsd.c", line 169,
function "pthread__add_specific"
Machine is amd64, 16 cores, KVM host. Among others "gunzip" fails.
Reverting this commit removes the failure.
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
On 29 May 2015, at 16:57, Emmanuel Dreyfus wrote:
> On Fri, May 29, 2015 at 04:26:33PM +0200, J. Hannken-Illjes wrote:
>> With this change some programs fail with
>>
>> assertion "pthread__tsd_destructors[key] != NULL" failed
>> file src/li
On 29 May 2015, at 16:26, J. Hannken-Illjes wrote:
> On 29 May 2015, at 09:37, Emmanuel Dreyfus wrote:
>
>> Module Name: src
>> Committed By:manu
>> Date:Fri May 29 07:37:32 UTC 2015
>>
>> Modified Files:
>> src/in
e date: 2015-06-01 08:50 UTC
The "official" atf runs for i386, amd64 and sparc show sporadic
failures for this test.
Does it fail on every run?
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
src/sys/miscfs/specfs: spec_vnops.c
>>
>> Log Message:
>> Revert previous, and explain why.
>
> Hrm, shouldn't it be &vp?
>
> (as it is, it's using the uvm_object lock pointer as the key)
Yes, just changed the key to the address of vp->v_specdev which i
tests:
> http://releng.netbsd.org/b5reports/i386/commits-2015.06.html#end
>
> The error message is:
> large_blk: panic: kernel diagnostic assertion "devvp->v_data == NULL"
> failed: file
> "/tmp/bracket/build/2015.06.30.06.42.06-i386/src/sys/ufs/mfs/mfs_vfsops.c",
> line 324
Should be fixed now. Passes ATF on i386.
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
Sure, will request pullups after one week (Jul 29).
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
On 27 Jul 2015, at 02:15, Erik Fair wrote:
> Pull up to netbsd-7 and netbsd-6?
>
> Erik
>
>
>> On Jul 22, 2015, at 01:33, Juergen
22 src/sys/dev/dkvar.h
> cvs rdiff -u -r1.91 -r1.92 src/sys/dev/ld.c
>
> Please note that diffs are not public domain; they are subject to the
> copyright notices on the relevant files.
>
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
On 27 Aug 2015, at 13:26, Michael van Elst wrote:
> On Thu, Aug 27, 2015 at 11:48:15AM +0200, J. Hannken-Illjes wrote:
>
>>disk_busy(&dksc->sc_dkdev);
>> + mutex_exit(&dksc->sc_iolock);
>>err
On 27 Aug 2015, at 13:26, Michael van Elst wrote:
> On Thu, Aug 27, 2015 at 11:48:15AM +0200, J. Hannken-Illjes wrote:
>
>>disk_busy(&dksc->sc_dkdev);
>> + mutex_exit(&dksc->sc_iolock);
>>err
tegy)(bp);
DEV_UNLOCK(d);
}
> and I am pretty sure you are not
> supposed to call biowait() on a biodone()'d buffer? And bdev_strategy
> should be made to return an error...
>
> christos
>
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
rc/external/gpl3/gcc.old/dist/gcc/varasm.c
> cvs rdiff -u -r1.3 -r1.4 \
>src/external/gpl3/gcc.old/dist/gcc/config/i386/i386.c
This change seems to break rump, number of failed tests on arch
amd64 goes from 19 to 1342, same problem on i386.
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
Trap: Fast Data Access MMU Miss
The attached hack works for me -- looks like we cannot use a mutex
so early ...
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
openfirm.diff
Description: Binary data
> On 05 Apr 2016, at 15:05, Martin Husemann wrote:
>
> On Tue, Apr 05, 2016 at 03:02:25PM +0200, J. Hannken-Illjes wrote:
>> The attached hack works for me -- looks like we cannot use a mutex
>> so early ...
>
> Or just skip the mutex_{enter,exit} while cold?
The f
eady protected by _VFS_VNODE_PRIVATE
and therefore invisible outside sys/kern/vfs_*.
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
T_INSERT_HEAD(&IN_IFADDR_HASH(ia->ia_addr.sin_addr.s_addr),
ia, ia_hash);
IN_ADDRHASH_WRITER_INSERT_HEAD(ia);
} else if (need_reinsert) {
This one looks wrong -- we enqueue "ia" twice.
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
eed before freeing them
makes the life of fsck harder.
- Running "brelse(bp, BC_INVAL)" doesn't look OK.
Please fix or revert soon.
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
*1 The stress is running the shell script attached. I usually get a
> On 07 Nov 2016, at 00:11, Jaromír Doleček wrote:
>
>> On Fri, Nov 04, 2016 at 04:44:10PM +0100, J. Hannken-Illjes wrote:
>>> - This change results in "panic: ffs_blkfree_common: freeing free block"
>>> if I put a file system under stress (*1).
>&g
> On 07 Nov 2016, at 11:54, Jaromír Doleček wrote:
>
> 2016-11-07 10:25 GMT+01:00 J. Hannken-Illjes :
>> The first part should not be necessary. After the loop we should have
>> "i == last" -- from a quick look "i < last" is impossible.
>
>
> On 07 Nov 2016, at 21:53, Jaromír Doleček wrote:
>
> 2016-11-07 12:11 GMT+01:00 J. Hannken-Illjes :
>> It gets used when a block of block pointers has been deallocated and
>> brelsed, that is NOT written to disk. If we allow the deallocation of
>> this block to
which allocation is wrong as its inode was in truncation.
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
wd_blkno == blk && wd->wd_len == len)
break;
}
KASSERT(wd != NULL);
wapbl_deallocation_free(wl, wd);
mutex_exit(&wl->wl_mtx);
}
> cvs rdiff -u -r1.19 -r1.20 src/sys/sys/wapbl.h
> cvs rdiff -u -r1.121 -r1.122 src/sys/ufs/ffs/ffs_inode.c
> cvs rdif
> On 11 Nov 2016, at 22:28, Jaromír Doleček wrote:
>
> 2016-11-11 11:52 GMT+01:00 J. Hannken-Illjes :
>> Returning a pointer to an arbitrary list element and using it
>> later is bad design. Would be better to define as:
>>
>> wapbl_unregister_deallocation(s
t is a MCLOSE device class?
> cvs rdiff -u -r1.34 -r1.35 src/sys/sys/audioio.h
> cvs rdiff -u -r1.146 -r1.147 src/sys/sys/conf.h
> cvs rdiff -u -r1.19 -r1.20 src/usr.bin/audio/ctl/audioctl.1
> cvs rdiff -u -r1.40 -r1.41 src/usr.bin/audio/ctl/ctl.c
>
> Please note that diffs a
> On 08 Dec 2016, at 15:41, Nathanial Sloss wrote:
>
> On Thu, 8 Dec 2016 23:57:51 J. Hannken-Illjes wrote:
>> +#defineD_MCLOSE0x0004
>>
>> What is a MCLOSE device class?
>>
>
> Normally character devices are don't execute the c
value should match userland,
either -1 on error or number of characters printed. The other kernel
printf variants should be changed too. It needs a kernel version bump.
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
getcwd_common().
>>
>> Should fix PR kern/53096 (netbsd-8 crash on heavy disk I/O)
>
> Pullup?
Already requested with ticket #702.
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
src/sys/kern: subr_pserialize.c
>
>> Log Message:
>> Function pserialize_perform() usually succeeds after two cross calls
>> so defer kpause() to iterations three and above.
>
>> Speeds up VOP_REVOKE() on /proc/XXX/status by a factor of ~12.
>
>> O
but missing from DESTDIR.
File wasn't installed ?
--
./usr/lib/i386/lua/5.3/netpgp.so
==== end of 1 missing files ==
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
1.192 -r1.193 src/usr.bin/make/job.c
>
> Please note that diffs are not public domain; they are subject to the
> copyright notices on the relevant files.
>
After this commit parallel builds take much longer. Building
amd64 release with -j16 for example goes from 45 to 380 minutes.
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
d64/lib/gcc/x86_64--netbsd/6.4.0/include-fixed
/build/devel/obj/dist.amd64/usr/include
Most includes come from "cddl/osnet/sys/sys" which looks wrong.
Suppose the search better starts with "-I${NETBSDSRCDIR}/sys" ...
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germa
domly
> fail (whenever fstrans_setstate() can actually take a long
> time without deadlock has to be determined, but I suspect it can).
>
> I think fstrans_setstate() should pause and retry in this case.
As Manuel said ...
This change is wrong, there is no upper time limit for a successf
unused field of struct lwp is reused.
For DIAGNOSTIC-only operations LWP specific data
(see kern/subr_lwp_specificdata.c) is a better choice.
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig
signature.asc
Description: Message signed with OpenPGP
> Add support for libprop as a host tool library.
>
>
> To generate a diff of this commit:
> cvs rdiff -u -r1.201 -r1.202 src/tools/Makefile
> cvs rdiff -u -r0 -r1.1 src/tools/libprop/Makefile
Looks like libprop needs pax.
Is the attached diff ok?
--
J. Hannken-Illje
. My Rev. 1.202 makefile has:
SUBDIR+= grep xz-lib libprop
.if ${TOOLS_BUILDRUMP} == "no"
SUBDIR += .WAIT texinfo \
.WAIT tic \
.WAIT pax \
.WAIT ${TOOLCHAIN_BITS} \
Here libprop needs pax, but pax gets built after it.
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
> On 7. May 2019, at 12:17, matthew green wrote:
>
>> Diff is NOT reversed.
>
> ah yes, i see. please commit.
Done.
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig (Germany)
c64/sparc64/db_trace.c:166:37:
note: each undeclared identifier is reported only once for each function it
appears in
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig
> On 22. May 2019, at 09:40, Martin Husemann wrote:
>
> Module Name: src
> Committed By: mar
Should be fixed with Rev. 1.4 of
src/external/cddl/osnet/dist/uts/common/fs/zfs/dmu_diff.c
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de - TU Braunschweig
> On 15. Oct 2019, at 00:13, Paul Goyette wrote:
>
> I'm now seeing a build error:
>
> # compile libzpool/dm
t; Please note that diffs are not public domain; they are subject to the
> copyright notices on the relevant files.
This looks wrong -- you should have added an weak alias to
sys/rump/librump/rumpkern/emul.c to handle the case where
librumpvfs is not present.
Please revert and fix correctly.
> On 1. Jun 2019, at 13:44, Robert Elz wrote:
>
>Date:Sat, 1 Jun 2019 10:34:41 +0200
>From: "J. Hannken-Illjes"
>Message-ID:
>
> | This looks wrong -- you should have added an weak alias to
> | sys/rump/librump/rumpke
gt;> defined(__DragonFly__) || defined(__OpenBSD__)
>> -# include
>> # else
>> # ifdef HAVE_ALLOCA_H
>> # include
>
> This looks like unintended change?
Fixed.
--
J. Hannken-Illjes - hann...@eis.cs.tu-bs.de <mailto:hann...@eis.cs.tu-bs.de> -
TU Braunschweig
signature.asc
Description: Message signed with OpenPGP
100 matches
Mail list logo
