bcc/SOURCES/bcc-0.8.0-print_log2_hist-check-and-skip-possible-paddings-215.patch

51 lines
2.0 KiB
Diff
Raw Normal View History

2019-08-01 18:34:48 +00:00
From 3f7b59660037c0d5dea785d115df25d9b95f07dc Mon Sep 17 00:00:00 2001
From: Xiaozhou Liu <liuxiaozhou@bytedance.com>
Date: Mon, 21 Jan 2019 11:23:42 +0800
Subject: [PATCH] print_log2_hist(): check and skip possible paddings (#2155)
Address issue 2154.
When a struct S is used as key to a BPF_HISTOGRAM, it is assumed that the second
member of S holds the slot. But when S is converted to python from bpf C,
a padding may be inserted as a second member. This breaks print_log2_hist().
root@debian:~/bcc/tools# ./softirqs.py -d
Tracing soft irq event time... Hit Ctrl-C to end.
^C
Traceback (most recent call last):
File "./softirqs.py", line 144, in <module>
dist.print_log2_hist(label, "softirq", section_print_fn=vec_to_name)
File "/usr/local/lib/python2.7/dist-packages/bcc/table.py", line 326, in print_log2_hist
vals[slot] = v.value
TypeError: list indices must be integers, not str
Fix it by skipping the possible padding. Future work would be fixing/working
around in the library where the padding is introduced.
---
src/python/bcc/table.py | 9 +++++++++
1 file changed, 9 insertions(+)
diff --git a/src/python/bcc/table.py b/src/python/bcc/table.py
index 6f598353..f6449de7 100644
--- a/src/python/bcc/table.py
+++ b/src/python/bcc/table.py
@@ -317,6 +317,15 @@ linear_index_max = 1025
tmp = {}
f1 = self.Key._fields_[0][0]
f2 = self.Key._fields_[1][0]
+
+ # The above code assumes that self.Key._fields_[1][0] holds the
+ # slot. But a padding member may have been inserted here, which
+ # breaks the assumption and leads to chaos.
+ # TODO: this is a quick fix. Fixing/working around in the BCC
+ # internal library is the right thing to do.
+ if f2 == '__pad_1' and len(self.Key._fields_) == 3:
+ f2 = self.Key._fields_[2][0]
+
for k, v in self.items():
bucket = getattr(k, f1)
if bucket_fn:
--
2.20.1