[parisc-linux] ntpdate broken

[parisc-linux] ntpdate broken

[Matthew Wilcox]

ntpdate isn't working on paer.  from the strace:

socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 4
fcntl(4, F_SETFL, O_RDONLY|O_NONBLOCK)  = 0
rt_sigaction(SIGALRM, {0x287a2, [], 0}, {SIG_DFL}, 8) = 0
setitimer(ITIMER_REAL, {it_interval={0, 200000}, it_value={0, 100000}}, NULL) = 0
setpriority(PRIO_PROCESS, 0, -12)       = 0
poll([{fd=4, events=POLLIN}], 1, 60000) = 0
poll( <unfinished ...>

anyone want to spend some time debugging this?  i'd bug the debian
maintainer, but i hear he's not very sympathetic towards hppa.

-- 
Revolutions do not require corporate support.

Re: [parisc-linux] ntpdate broken

[Richard Hirst]

On Wed, Dec 05, 2001 at 04:18:53AM +0000, Matthew Wilcox wrote:
> 
> ntpdate isn't working on paer.  from the strace:
> 
> socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 4
> fcntl(4, F_SETFL, O_RDONLY|O_NONBLOCK)  = 0
> rt_sigaction(SIGALRM, {0x287a2, [], 0}, {SIG_DFL}, 8) = 0
> setitimer(ITIMER_REAL, {it_interval={0, 200000}, it_value={0, 100000}}, NULL) = 0
> setpriority(PRIO_PROCESS, 0, -12)       = 0
> poll([{fd=4, events=POLLIN}], 1, 60000) = 0
> poll( <unfinished ...>
> 
> anyone want to spend some time debugging this?  i'd bug the debian
> maintainer, but i hear he's not very sympathetic towards hppa.

That poll call should have been interrupted by SIGALRM after about 100ms.
time_t is long, so I guess we need a wrapper for setitimer()/getitimer().

Richard

Re: [parisc-linux] ntpdate broken

[Carlos O'Donell Jr.]

> > 
> > ntpdate isn't working on paer.  from the strace:
> > 
> > socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 4
> > fcntl(4, F_SETFL, O_RDONLY|O_NONBLOCK)  = 0
> > rt_sigaction(SIGALRM, {0x287a2, [], 0}, {SIG_DFL}, 8) = 0
> > setitimer(ITIMER_REAL, {it_interval={0, 200000}, it_value={0, 100000}}, NULL) = 0
> > setpriority(PRIO_PROCESS, 0, -12)       = 0
> > poll([{fd=4, events=POLLIN}], 1, 60000) = 0
> > poll( <unfinished ...>
> > 
> > anyone want to spend some time debugging this?  i'd bug the debian
> > maintainer, but i hear he's not very sympathetic towards hppa.
> 
> That poll call should have been interrupted by SIGALRM after about 100ms.
> time_t is long, so I guess we need a wrapper for setitimer()/getitimer().
> 
> Richard

Could there be other long issues lurking?
A hidden assumption somewhere?

I guess my questions would be: 
Are wide/narrow wrapper issues ferreted out one by one?

c.