Clock.cxx: some bug fixes

[Jonathan Larmour]

Nick Garnett wrote:
> The following patch fixes a couple of bugs that have popped up on the
> discussion list recently. The rem_alarm() fix is straightforward.
> 
> The tick() fix is a total rewrite of a section of code. Before
> committing it I would like some others to look it over and check that
> it does what it is supposed to. Since CVS diff has made the new code
> hard to read in the patch, I have appended a clean copy to the end of
> this message.

Perhaps the thing to do is write a test (or expand an existing one) that 
pokes it in the right way to cause the problems it's trying to fix, i.e. 
tweaking the alarms from within an alarm handler.

> If I hear nothing in a few days, I'll go ahead and commit it.

I had already applied Thomas' patch (in a very slightly different way) at 
almost exactly the same time you sent this mail, sorry! The only 
difference (as I indicated in my mail) was to lock the scheduler _after_ 
the assign to alarm_list_ptr in the single list case. See the attached 
patch FAOD.

I can't help but think that the extra checking for changes i.e. next_next, 
next_prev and head, can be a config option given this is a well-trodden 
piece of code.

Well you did imply you wanted comments ;).

Jifl
-- 
eCosCentric    http://www.eCosCentric.com/    The eCos and RedBoot experts
--[ "You can complain because roses have thorns, or you ]--
--[  can rejoice because thorns have roses." -Lincoln   ]-- Opinions==mine
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: alarm.patch
URL: <http://sourceware.org/pipermail/ecos-patches/attachments/20030626/069359a0/attachment.ksh>