Demonstrate the bug where an aligned write can be lost if it races
between the read and write of a RMW unaligned write.
---
Sending this out for review of the test; it fails (which it is
supposed to as long as I don't fix the blocksize filter), but I hope
it starts passing once I also patch the
Prior to this patch, the following fails, but at least seems to give a
sensible error:
$ nbdsh -c 'nbd.Buffer.from_bytearray(b"1"*8)'
Traceback (most recent call last):
File "/usr/lib64/python3.10/runpy.py", line 196, in _run_module_as_main
return _run_code(code, main_globals, None,
File
On May 25 2022, Eric Blake wrote:
> On Tue, May 24, 2022 at 08:45:02PM +0100, Nikolaus Rath wrote:
>> > However, you are worried that a third possibility occurs:
>> >
>> > T2 sees that it needs to do RMW, grabs the lock, and reads 0x00-0x0f
>> > for the unaligned head (it only needs 0x00-0x03,
On Wed, 25 May 2022 at 18:34, Richard W.M. Jones wrote:
>
> On Wed, May 25, 2022 at 05:13:53PM +0100, Peter Maydell wrote:
> > On Wed, 25 May 2022 at 16:07, Laszlo Ersek wrote:
> > >
> > > + Drew & Peter
> > >
> > > On 05/25/22 15:30, Daniel P. Berrangé wrote:
> > > - The patch seems to do what
On Tue, May 24, 2022 at 08:45:02PM +0100, Nikolaus Rath wrote:
> > However, you are worried that a third possibility occurs:
> >
> > T2 sees that it needs to do RMW, grabs the lock, and reads 0x00-0x0f
> > for the unaligned head (it only needs 0x00-0x03, but we have to read a
> > block at a time),
On Wed, May 25, 2022 at 05:13:53PM +0100, Peter Maydell wrote:
> On Wed, 25 May 2022 at 16:07, Laszlo Ersek wrote:
> >
> > + Drew & Peter
> >
> > On 05/25/22 15:30, Daniel P. Berrangé wrote:
> > - The patch seems to do what it says in the commit message.
> >
> > - QEMU commit bab52d4bba3f
On Wed, 25 May 2022 at 16:07, Laszlo Ersek wrote:
>
> + Drew & Peter
>
> On 05/25/22 15:30, Daniel P. Berrangé wrote:
> - The patch seems to do what it says in the commit message.
>
> - QEMU commit bab52d4bba3f ("target/arm: Add "-cpu max" support",
> 2018-03-09) confirms what the commit
We previously didn't bother to check the return values from any librpm
calls. In some cases where possibly the RPM database is faulty, this
caused us to return a zero-length list of installed applications (but
no error indication).
One way to reproduce this is given below. Note this reproducer
In libguestfs we didn't bother to check the return values from any
librpm calls. In some cases where possibly the RPM database is
faulty, this caused us to return a zero-length list of installed
applications (but no error indication). Libguestfs has subsequently
been fixed so now it returns an
+ Drew & Peter
On 05/25/22 15:30, Daniel P. Berrangé wrote:
> The current logic for selecting CPU model to use with the appliance
> selects 'max' for all architectures except for ppc64le and aarch64.
>
> On aarch64, it selects 'host' if KVM is available which is identical
> to 'max'. For TCG it
The current logic for selecting CPU model to use with the appliance
selects 'max' for all architectures except for ppc64le and aarch64.
On aarch64, it selects 'host' if KVM is available which is identical
to 'max'. For TCG it selects 'cortex-a57'. The 'max' model is a
superset of 'cortex-a57', so
On 05/25/22 11:36, Richard W.M. Jones wrote:
>
> I have tested both patches this time with the three affected tools
> (virt-builder, virt-customize & virt-sysprep) and the --selinux-relabel
> option is now present and ignored.
>
> For the series:
> Reviewed-by: Richard W.M. Jones
>
> Thanks
I've updated the HTML man pages on the website:
https://libguestfs.org/nbdkit.1.html
Note the synopsis section of that first page is wrong, but I kind of
knew that already. That particular section is generated by the
podwrapper --verbatim option, unlike all the others.
Rich.
--
Richard
I have tested both patches this time with the three affected tools
(virt-builder, virt-customize & virt-sysprep) and the --selinux-relabel
option is now present and ignored.
For the series:
Reviewed-by: Richard W.M. Jones
Thanks for fixing this, I'll get it into C9S soon.
Rich.
--
Richard
Thanks - the series was pushed in:
b2612d2c..43b341b0 master -> master
Rich.
--
Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones
Read my programming and virtualization blog: http://rwmj.wordpress.com
Fedora Windows cross-compiler. Compile Windows programs,
Refresh the generated mlcustomize files after libguestfs patch
'generator/customize: reintroduce "--selinux-relabel" as a compat option'.
Fixes: 3420d5f14f6d0a713a40d1c2ac56677eba3c5ac8
Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=2089748
Signed-off-by: Laszlo Ersek
---
Removing "--selinux-relabel" in commit 2f6a27f1077d ("generator/customize:
invert SELinux relabeling default", 2022-05-11) breaks existing scripts
that invoke virt-customize and/or virt-sysprep with that option. Restore
the option, with no functionality tied to it.
Fixes:
On 05/25/22 07:24, Laszlo Ersek wrote:
> On 05/24/22 15:59, Laszlo Ersek wrote:
>> On 05/24/22 12:04, Richard W.M. Jones wrote:
>>> So I didn't realise that these commits break --selinux-relabel, eg:
>>>
>>> $ rpm -q guestfs-tools
>>> guestfs-tools-1.49.1-1.fc37.x86_64
>>> $ virt-builder fedora-36
18 matches
Mail list logo