2005-04-08 14:38:33

by Tom Rini

[permalink] [raw]
Subject: [PATCH 2.6.12-rc2]: ppc32: Fix a problem with NTP on !(chrp||gemini)

The following problem was found by Giovambattista Pulcini
<[email protected]>, who also provided a partial patch, and this has
been verified by our time guru Gabriel Paubert <[email protected]>.

The problem is that in do_settimeofday() we always set time_state to
TIME_ERROR and except on two platforms, never re-set it. This meant
that ntp_gettime() and ntp_adjtime() always returned TIME_ERROR,
incorrectly. Based on Gabriel's analysis, time_state is used for
leap-second processing, and ppc shouldn't be mucking with it.

From: Giovambattista Pulcini <[email protected]>
Signed-off-by: Tom Rini <[email protected]>

--- 1.37/arch/ppc/kernel/time.c 2005-01-20 22:02:08 -07:00
+++ edited/arch/ppc/kernel/time.c 2005-04-08 07:30:46 -07:00
@@ -272,7 +272,6 @@

time_adjust = 0; /* stop active adjtime() */
time_status |= STA_UNSYNC;
- time_state = TIME_ERROR; /* p. 24, (a) */
time_maxerror = NTP_PHASE_LIMIT;
time_esterror = NTP_PHASE_LIMIT;
write_sequnlock_irqrestore(&xtime_lock, flags);
--- 1.11/arch/ppc/platforms/chrp_time.c 2005-01-25 14:50:14 -07:00
+++ edited/arch/ppc/platforms/chrp_time.c 2005-04-08 07:30:53 -07:00
@@ -115,8 +115,6 @@
chrp_cmos_clock_write(save_control, RTC_CONTROL);
chrp_cmos_clock_write(save_freq_select, RTC_FREQ_SELECT);

- if ( (time_state == TIME_ERROR) || (time_state == TIME_BAD) )
- time_state = TIME_OK;
spin_unlock(&rtc_lock);
return 0;
}
--- 1.17/arch/ppc/platforms/gemini_setup.c 2005-03-13 16:29:43 -07:00
+++ edited/arch/ppc/platforms/gemini_setup.c 2005-04-08 07:30:56 -07:00
@@ -433,9 +433,6 @@
/* done writing */
gemini_rtc_write(reg, M48T35_RTC_CONTROL);

- if ((time_state == TIME_ERROR) || (time_state == TIME_BAD))
- time_state = TIME_OK;
-
return 0;
}

--
Tom Rini
http://gate.crashing.org/~trini/