blob: 64a1d5326d75794a5072709d988bdaeb447a88c3 (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
|
From 324ac9933b9b4d2f477942c8b1f1c54241933a76 Mon Sep 17 00:00:00 2001
From: Nikolaus Rath <Nikolaus@rath.org>
Date: Tue, 23 Feb 2016 15:44:51 -0800
Subject: Estimate system clock granularity more conservatively
Origin: debian
Forwarded: no
Patch-Name: clock-granularity.diff
Some unit tests check that file access time stamps are updated
correctly. For such a test to succeed, it's necessarily that the time
between two subsequent test-accesses can actually be resolved. The
heuristic to determine this granularity doesn't seem to work well on
e.g. the i386 buildd, so we make it much more conservative. This means
the test suite runs longer, but otherwise has no ill effects.
---
tests/common.py | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tests/common.py b/tests/common.py
index 93736..af5ed 100644
--- a/tests/common.py
+++ b/tests/common.py
@@ -31,7 +31,7 @@ def get_clock_granularity():
stamp2 = time.time()
resolution = min(resolution, stamp2 - stamp1)
time.sleep(0.01)
- return resolution
+ return max(1, 10 * resolution)
CLOCK_GRANULARITY = get_clock_granularity()
# When testing, we want to make sure that we don't sleep for too short a time
|
