28 Sep
2005
28 Sep
'05
9:49 a.m.
On Tue, 27 Sep 2005, Mark Knecht wrote:
LD .tmp_vmlinux1
drivers/built-in.o(.text+0x24acc): In function `acpi_processor_idle':
: undefined reference to `tsc_c3_compensate'
make: *** [.tmp_vmlinux1] Error 1
lightning linux #
The 2.6.13 kernel builds fine before I apply the patch but fails afterward.
Neither vanilla 2.6.13 nor 2.6.13.1 has any reference to tsc_c3_compensate
but I googled this from LKM (http://seclists.org/lists/linux-kernel/2005/Aug/2889.html)
+/* Code to mark and check if the TSC is unstable
+ * due to cpufreq or due to unsynced TSCs
+ */
+static int tsc_unstable;
+int check_tsc_unstable(void)
+{
+return tsc_unstable;
+}
+
+void mark_tsc_unstable(void)
+{
+tsc_unstable = 1;
+}
+
+/* Code to compensate for C3 stalls */
+static u64 tsc_c3_offset;
+void tsc_c3_compensate(unsigned long usecs)
+{
+u64 cycles = (usecs * tsc_khz)/1000;
+tsc_c3_offset += cycles;
+}
+
+u64 tsc_read_c3_time(void)
+{
+return tsc_c3_offset;
+}
+
Seems to be 64 related, but not yet in mainline. Perhaps Ingo
mistakenly included or referred it in his patch
NOTE: I have tried other realtime kernels like
ck-sources but so far I
am having xrun problems. Possibly there is someone here successfully
using this motherboard? If so please contact me on-line or off with
info on your setup.
Have you tried the rtlimits/patched PAM solution?
On my 32bit AthlonXP 1200+ with Audiophile 2496 I have no problems.
- Peder