> Why would I want to nuke them? Shouldnt I be looking in them, for logs if
> nothing else?
YMMV and such. If it works, I'm generally done with them.
They get out of date. I hate working with out-of-date stuff - too much
opportunity for confusion.
I often grep the whole world to find the us
Yo Hal!
On Tue, 05 Dec 2017 12:48:05 -0800
Hal Murray wrote:
> > Well, not exactly true, git status complains about:
> > test-all/
> > test-classic/
> > test-default/
> > test-doc/
> > test-minimal/
> > They should prolly be in .gitignore
>
> I would be happy to add them,
> Well, not exactly true, git status complains about:
> test-all/
> test-classic/
> test-default/
> test-doc/
> test-minimal/
> They should prolly be in .gitignore
I would be happy to add them, but just as happy to leave them out so git
status reminds me to nuke th
Yo Hal!
On Tue, 05 Dec 2017 11:56:15 -0800
Hal Murray wrote:
> > I try my build script first after clone, and option-tester.sh first
> > after clone. I no longer see the issue. Weird...
>
> Can you diff the directories?
Nope, I blew off my old copy. But before I did, git status showed no
> I try my build script first after clone, and option-tester.sh first after
> clone. I no longer see the issue. Weird...
Can you diff the directories?
Another possibility is that ccache got confused.
--
These are my opinions. I hate spam.
___
Yo Hal!
On Mon, 04 Dec 2017 20:10:09 -0800
Hal Murray wrote:
> > I get exactly the same results. ???
> > Then my build fails:
>
> Just to make sure...
I'm all ears, this is a weird one, and hopefully related to other
peoples odd issues with DEBUG.
> Any chance you have local edits doing som
> I get exactly the same results. ???
> Then my build fails:
Just to make sure...
Any chance you have local edits doing something strange?
Can you try it on a fresh git clone?
Does gcc look in any environment variables?
Another possibility is something like:
grep "#define" . -r | grep DEBUG
Yo Hal!
On Mon, 04 Dec 2017 19:38:00 -0800
Hal Murray wrote:
> > I can confirm that debug is being set.
>
> This is pretty weird. I wonder where it is being set, and why it
> isn't also set in the other modules.
>
> > I don not see anything in there to turn on DEBUG.
> > Is there a verbose
> I can confirm that debug is being set.
This is pretty weird. I wonder where it is being set, and why it isn't also
set in the other modules.
> I don not see anything in there to turn on DEBUG.
> Is there a verbose option for waf to show the gcc commands it is running?
waf -v is verbose mode.
g...@rellim.com said:
> Since I do not have NTP_DEBUG_LISTS defined, the extern is never pulled in.
> Not the first time thee DBUG stuff caused issues for me.
Look in include/ntp_lists.h
if DEBUG exists, it defines NTP_DEBUG_LISTS
and sets up a couple of list macros to call check_gen_fifo_consi
Yo Hal!
On Mon, 04 Dec 2017 13:31:43 -0800
Hal Murray wrote:
> devel@ntpsec.org said:
> >> I'd like to encourage everybody to use it when making
> > "interesting" changes.
> > Mostly failed for me. See attached.
>
> [93/93] Linking test-default/main/ntptime/ntptime
> ntpd/ntp_config.c.3.o
devel@ntpsec.org said:
>> I'd like to encourage everybody to use it when making
> "interesting" changes.
> Mostly failed for me. See attached.
[93/93] Linking test-default/main/ntptime/ntptime
ntpd/ntp_config.c.3.o: In function `destroy_attr_val_fifo':
ntp_config.c:(.text+0x33): undefined refer
Yo Hal!
On Mon, 04 Dec 2017 03:20:35 -0800
Hal Murray via devel wrote:
> I'd like to encourage everybody to use it when making "interesting"
> changes.
Mostly failed for me. See attached.
RGDS
GARY
---
Gary E. Miller Rell
I'd like to encourage everybody to use it when making "interesting" changes.
The idea is to at least try all the configuration options. It will catch
simple quirks before you do the push. (Better is to try it before you do the
commit so we avoid broken landing spots when doing a bisect.)
The
14 matches
Mail list logo
