On Thu, Jan 29, 2009 at 10:48:36PM -0500, Dave Allan wrote:
Jim Meyering wrote:
Finally, I moved a couple variable declarations
down (C99-style) to their points of first use.
I take it that the C++/C99 is the recommended style for all libvirt
code? I generally haven't coded this way in
On Fri, Jan 30, 2009 at 12:07:31AM +0100, Jim Meyering wrote:
Here's your rebased and adjusted patch:
From ce4f15853e119d6d976a5d29917f62f577e8ec9e Mon Sep 17 00:00:00 2001
From: Jim Meyering meyer...@redhat.com
Date: Thu, 29 Jan 2009 22:50:36 +0100
Subject: [PATCH] allow disk cache mode
Daniel P. Berrange berra...@redhat.com wrote:
...
Further up in this code its already adding cache=off for shared disks.
Probably introduced by me in the merge.
There's also quite a few code style issues, not following conventions
of the surrounding code.
Here's a patch which addresses all
On Fri, Jan 30, 2009 at 03:27:55PM +0100, Jim Meyering wrote:
Daniel P. Berrange berra...@redhat.com wrote:
...
Further up in this code its already adding cache=off for shared disks.
Probably introduced by me in the merge.
There's also quite a few code style issues, not following
On Fri, Jan 30, 2009 at 02:41:18PM +, Daniel P. Berrange wrote:
On Fri, Jan 30, 2009 at 03:27:55PM +0100, Jim Meyering wrote:
Daniel P. Berrange berra...@redhat.com wrote:
...
Further up in this code its already adding cache=off for shared disks.
Probably introduced by me in the
Daniel P. Berrange berra...@redhat.com wrote:
Here's the patch including RNG schema update
Daniel
diff -r 2ff2ff7734c2 docs/schemas/domain.rng
--- a/docs/schemas/domain.rng Fri Jan 30 11:01:52 2009 +
+++ b/docs/schemas/domain.rng Fri Jan 30 14:47:11 2009 +
@@ -426,16 +426,43 @@
Jim Meyering wrote:
Hi John,
I tried to apply that, but failed miserably,
since all of the following was recently redone to
use virBufferVSprintf rather than snprintf.
Yea I suspected the code was likely seeing some
motion. Thanks for bringing it forward.
And it's
a good thing, because with
john cooper john.coo...@redhat.com wrote:
Daniel P. Berrange wrote:
If you loook at src/qemu_conf.c, you'll find a nice method called
qemudExtractVersionInfo, which runs 'qemu -help' and checks for
certain interesting command line arguments :-)
That problem does seem to be crying for some
Jim Meyering wrote:
Finally, I moved a couple variable declarations
down (C99-style) to their points of first use.
I take it that the C++/C99 is the recommended style for all libvirt
code? I generally haven't coded this way in the past (in fact I usually
compile with
Daniel P. Berrange wrote:
If you loook at src/qemu_conf.c, you'll find a nice method called
qemudExtractVersionInfo, which runs 'qemu -help' and checks for
certain interesting command line arguments :-)
That problem does seem to be crying for some type
of structured interface to avoid subtle
On Mon, Jan 19, 2009 at 10:06:47PM -0500, john cooper wrote:
There have been two differnet syntaxes supported in QEMU for caching,
so we need to detect this and switch between them. Originally there
was just
cache=on cache=off (writethrough and no caching)
Now it supports
On Fri, Jan 16, 2009 at 03:46:01PM -0500, john cooper wrote:
We have found certain application scenarios where
overriding of the default qemu host cache mode
provides a substantial improvement in guest
performance. In particular, disabling host caching
of the file/dev backing a guest drive.
Daniel P. Berrange wrote:
A couple of extra things needed
- Addition to tests/qemuxml2argvtest.c to validate the XML to
struct to QEMU ARGV conversion.
- Addition to tests/qemuxml2xmltest.c to validate XML to
struct to XML round-trip conversions.
- Addition to the docs/libvirt.rng
We have found certain application scenarios where
overriding of the default qemu host cache mode
provides a substantial improvement in guest
performance. In particular, disabling host caching
of the file/dev backing a guest drive. A summary
of performance metrics may be found below.
Attached
14 matches
Mail list logo