[haiku-development] Re: Alpha3 : switch to bugfixing mode today

  • From: Stephan Aßmus <superstippi@xxxxxx>
  • To: haiku-development@xxxxxxxxxxxxx
  • Date: Mon, 16 May 2011 10:00:31 +0200

On 15.05.2011 23:25, Sean Collins wrote:
Adrien Destugues wrote:
Le 15/05/2011 18:57, Stephan Aßmus a écrit :
On 14.05.2011 10:56, Adrien Destugues wrote:
From the 30 tickets in "alpha3" milestone two weeks ago, there is only
one left :
https://dev.haiku-os.org/query?status=assigned&status=in-progress&status=reopened&status=new&group=status&milestone=R1%2Falpha3

One week from now the alpha branch will be created. It's time to fix
that remaining bug, and start a test run of nightlies to see if
everything looks ok. We may get some more bugs to fix after test
reports.

What happened to the other 29 tickets? (I am not on the ticket list
anymore.) Were those all not suitable for blocking Alpha 3 in the
end, or did Michael's work on IO-APIC resolve most of them?

There was a mix of :
* Tickets moved to Beta 1 or R1, because they were actually
enhancement requests
* Tickets merged because they were multiple instance of the same issue
("not being able to boot from USB"). These are hard to debug since the
problem is very early in the boot sequence. Starseed is investigating
on them right now. It seems to be a problem in the stage1 bootloader.
* Some were also about problems with CD images, these were fixed by an
update of BOM FreeBSD and mkisofs.
* I think some were also related to problems with ported software,
scottmc may tell you more on that. There was a set of updates to fix
them.
* And of course IO-APIC handling has also helped in some cases, even
if it doesn't work for everyone yet.

Thanks Adrien and Scott for the information!

Enabling by default the IO-APIC for the debug week might be very
enlightening for data collection purposes. See what types of problems
and which hardware has problems. Would be nice to see this big bug
fixed. Might get a flood of tickets, might not get a flood of tickets ??

Could be a good idea. I checked with r41462 (but only on my main computer) and IO-APIC worked fine on it (checked the syslog, it was used). I only looked at one ticket more closely where there was a chance of IO APIC doing some good, but it wouldn't resolve the issue, yet. Are there any reports that the new IO APIC stuff helped anyone get Haiku finally booting on there hardware?

Best regards,
-Stephan

Other related posts: