Am 17.05.2011 um 22:39 schrieb Blue Swirl:
On Tue, May 17, 2011 at 5:47 AM, David Ahern <daah...@cisco.com>
wrote:
On 05/16/11 13:56, Gerd Hoffmann wrote:
This patch finally merges the EHCI host adapter aka USB 2.0 support.
Based on git://git.kiszka.org/qemu.git ehci
Changes:
- Adapt to recent changes in the usb subsystem.
- Don't create device automagically, use -device instead.
- Add quickstart text file, see docs/usb2.txt.
- A bunch of codestyle fixups.
- Add authors+contributers list.
- Zap EHCI_NOMICROFRAMES, qemu can't handle a 8 kHz
wakeup rate anyway.
- A few bug fixes.
Signed-off-by: Gerd Hoffmann <kra...@redhat.com>
As someone who spent a significant amount of time working on the EHCI
code last year I am absolutely not ok with this. The entire
contribution
history for EHCI lost - and for no reason. The inclusion of EHCI into
qemu can be done in such a way as to maintain the history.
The development history (implement a feature, fix bugs, implement
another, fix bugs etc) is not interesting and pulling it would make
bisection and other bugfixing more difficult. The patches should not
add known broken features and fix them next.
Instead, patches should add one simple, bug free feature at a time.
This also makes it possible for some of them to be committed before
others, while others may get rejected and may need major rework. The
patches for EHCI are just like other patches.
Well, if David and Jan contributed to it, just as for other patches it
would've been good practice to at least Cc: them. That's a way to
carry a contribution history even if patches are reworked.
Reminds me of Jonathan Corbet's FOSDEM keynote about kernel
contributions: In the end it's not just about fixing a problem; if you
liberally recycle other people's patches and commit them as your own,
you risk losing contributors.
Andreas
