X-Git-Url: https://pintos-os.org/cgi-bin/gitweb.cgi?p=pintos-anon;a=blobdiff_plain;f=src%2Ftests%2Fthreads%2Fmlfqs-load-avg.c;h=50e83e276406714b03ab8d199a16259ee574ffbf;hp=8ea4dd14a9243638f16dec6ffefde2b753a1f6a1;hb=d4def0ac6ef072cf0cf8a968784bc21371bd4d58;hpb=f0563386a8c672a1460473b0b4b175e74eaf5c5c

diff --git a/src/tests/threads/mlfqs-load-avg.c b/src/tests/threads/mlfqs-load-avg.c
index 8ea4dd1..50e83e2 100644
--- a/src/tests/threads/mlfqs-load-avg.c
+++ b/src/tests/threads/mlfqs-load-avg.c
@@ -4,7 +4,17 @@
    seconds, starting 10 seconds in, the main thread prints the
    load average.
 
-   The expected output is this (some margin of error is allowed):
+   The expected output is listed below.  Some margin of error is
+   allowed.
+
+   If your implementation fails this test but passes most other
+   tests, then consider whether you are doing too much work in
+   the timer interrupt.  If the timer interrupt handler takes too
+   long, then the test's main thread will not have enough time to
+   do its own work (printing a message) and go back to sleep
+   before the next tick arrives.  Then the main thread will be
+   ready, instead of sleeping, when the tick arrives,
+   artificially driving up the load average.
 
    After 0 seconds, load average=0.00.
    After 2 seconds, load average=0.05.