summaryrefslogtreecommitdiffstats
path: root/meta/lib
diff options
context:
space:
mode:
authorRichard Purdie <richard.purdie@linuxfoundation.org>2013-03-14 17:26:20 +0000
committerRichard Purdie <richard.purdie@linuxfoundation.org>2013-03-18 21:26:40 +0000
commitadd11fa1abfc51f1a793c84f02cd7d85d793fe14 (patch)
treee58176a867d272a57d91f61e2f954fc0249d007e /meta/lib
parentcec0102647e3f2c93de5125c1de2436b4b787bdd (diff)
downloadpoky-add11fa1abfc51f1a793c84f02cd7d85d793fe14.tar.gz
package: Add cachedpath optimisation
Currently, various standard library operations like os.walk(), os.path.isdir() and os.path.islink() each call stat or lstat which involves a syscall into the kernel. There is no caching since they could conceivably have changed on disk. The result is that for something like the do_package task of the kernel we're spending over two minutes making 868,000 individual stat calls for 23,000 files. This is suboptimal. This patch adds lib/oe/cachedpath.py which are a set of replacement functions for these operations which use cached stat data rather than hitting the kernel each time. It gives a nice performance improvement halving the build time of the kernel do_package. (From OE-Core rev: 556dee0c4d6d8a87c0cddbd2f60fe5917d009f18) Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Diffstat (limited to 'meta/lib')
-rw-r--r--meta/lib/oe/cachedpath.py235
1 files changed, 235 insertions, 0 deletions
diff --git a/meta/lib/oe/cachedpath.py b/meta/lib/oe/cachedpath.py
new file mode 100644
index 0000000000..e350c8a70e
--- /dev/null
+++ b/meta/lib/oe/cachedpath.py
@@ -0,0 +1,235 @@
1#
2# Based on standard python library functions but avoid
3# repeated stat calls. Its assumed the files will not change from under us
4# so we can cache stat calls.
5#
6
7import os
8import errno
9import stat as statmod
10
11class CachedPath(object):
12 def __init__(self):
13 self.statcache = {}
14 self.lstatcache = {}
15 self.normpathcache = {}
16 return
17
18 def updatecache(self, x):
19 x = self.normpath(x)
20 if x in self.statcache:
21 del self.statcache[x]
22 if x in self.lstatcache:
23 del self.lstatcache[x]
24
25 def normpath(self, path):
26 if path in self.normpathcache:
27 return self.normpathcache[path]
28 newpath = os.path.normpath(path)
29 self.normpathcache[path] = newpath
30 return newpath
31
32 def _callstat(self, path):
33 if path in self.statcache:
34 return self.statcache[path]
35 try:
36 st = os.stat(path)
37 self.statcache[path] = st
38 return st
39 except os.error:
40 self.statcache[path] = False
41 return False
42
43 # We might as well call lstat and then only
44 # call stat as well in the symbolic link case
45 # since this turns out to be much more optimal
46 # in real world usage of this cache
47 def callstat(self, path):
48 path = self.normpath(path)
49 self.calllstat(path)
50 return self.statcache[path]
51
52 def calllstat(self, path):
53 path = self.normpath(path)
54 if path in self.lstatcache:
55 return self.lstatcache[path]
56 #bb.error("LStatpath:" + path)
57 try:
58 lst = os.lstat(path)
59 self.lstatcache[path] = lst
60 if not statmod.S_ISLNK(lst.st_mode):
61 self.statcache[path] = lst
62 else:
63 self._callstat(path)
64 return lst
65 except (os.error, AttributeError):
66 self.lstatcache[path] = False
67 self.statcache[path] = False
68 return False
69
70 # This follows symbolic links, so both islink() and isdir() can be true
71 # for the same path ono systems that support symlinks
72 def isfile(self, path):
73 """Test whether a path is a regular file"""
74 st = self.callstat(path)
75 if not st:
76 return False
77 return statmod.S_ISREG(st.st_mode)
78
79 # Is a path a directory?
80 # This follows symbolic links, so both islink() and isdir()
81 # can be true for the same path on systems that support symlinks
82 def isdir(self, s):
83 """Return true if the pathname refers to an existing directory."""
84 st = self.callstat(s)
85 if not st:
86 return False
87 return statmod.S_ISDIR(st.st_mode)
88
89 def islink(self, path):
90 """Test whether a path is a symbolic link"""
91 st = self.calllstat(path)
92 if not st:
93 return False
94 return statmod.S_ISLNK(st.st_mode)
95
96 # Does a path exist?
97 # This is false for dangling symbolic links on systems that support them.
98 def exists(self, path):
99 """Test whether a path exists. Returns False for broken symbolic links"""
100 if self.callstat(path):
101 return True
102 return False
103
104 def lexists(self, path):
105 """Test whether a path exists. Returns True for broken symbolic links"""
106 if self.calllstat(path):
107 return True
108 return False
109
110 def stat(self, path):
111 return self.callstat(path)
112
113 def lstat(self, path):
114 return self.calllstat(path)
115
116 def walk(self, top, topdown=True, onerror=None, followlinks=False):
117 # Matches os.walk, not os.path.walk()
118
119 # We may not have read permission for top, in which case we can't
120 # get a list of the files the directory contains. os.path.walk
121 # always suppressed the exception then, rather than blow up for a
122 # minor reason when (say) a thousand readable directories are still
123 # left to visit. That logic is copied here.
124 try:
125 # Note that listdir and error are globals in this module due
126 # to earlier import-*.
127 names = os.listdir(top)
128 except error, err:
129 if onerror is not None:
130 onerror(err)
131 return
132
133 dirs, nondirs = [], []
134 for name in names:
135 if self.isdir(os.path.join(top, name)):
136 dirs.append(name)
137 else:
138 nondirs.append(name)
139
140 if topdown:
141 yield top, dirs, nondirs
142 for name in dirs:
143 new_path = os.path.join(top, name)
144 if followlinks or not self.islink(new_path):
145 for x in self.walk(new_path, topdown, onerror, followlinks):
146 yield x
147 if not topdown:
148 yield top, dirs, nondirs
149
150 ## realpath() related functions
151 def __is_path_below(self, file, root):
152 return (file + os.path.sep).startswith(root)
153
154 def __realpath_rel(self, start, rel_path, root, loop_cnt, assume_dir):
155 """Calculates real path of symlink 'start' + 'rel_path' below
156 'root'; no part of 'start' below 'root' must contain symlinks. """
157 have_dir = True
158
159 for d in rel_path.split(os.path.sep):
160 if not have_dir and not assume_dir:
161 raise OSError(errno.ENOENT, "no such directory %s" % start)
162
163 if d == os.path.pardir: # '..'
164 if len(start) >= len(root):
165 # do not follow '..' before root
166 start = os.path.dirname(start)
167 else:
168 # emit warning?
169 pass
170 else:
171 (start, have_dir) = self.__realpath(os.path.join(start, d),
172 root, loop_cnt, assume_dir)
173
174 assert(self.__is_path_below(start, root))
175
176 return start
177
178 def __realpath(self, file, root, loop_cnt, assume_dir):
179 while self.islink(file) and len(file) >= len(root):
180 if loop_cnt == 0:
181 raise OSError(errno.ELOOP, file)
182
183 loop_cnt -= 1
184 target = os.path.normpath(os.readlink(file))
185
186 if not os.path.isabs(target):
187 tdir = os.path.dirname(file)
188 assert(self.__is_path_below(tdir, root))
189 else:
190 tdir = root
191
192 file = self.__realpath_rel(tdir, target, root, loop_cnt, assume_dir)
193
194 try:
195 is_dir = self.isdir(file)
196 except:
197 is_dir = False
198
199 return (file, is_dir)
200
201 def realpath(self, file, root, use_physdir = True, loop_cnt = 100, assume_dir = False):
202 """ Returns the canonical path of 'file' with assuming a
203 toplevel 'root' directory. When 'use_physdir' is set, all
204 preceding path components of 'file' will be resolved first;
205 this flag should be set unless it is guaranteed that there is
206 no symlink in the path. When 'assume_dir' is not set, missing
207 path components will raise an ENOENT error"""
208
209 root = os.path.normpath(root)
210 file = os.path.normpath(file)
211
212 if not root.endswith(os.path.sep):
213 # letting root end with '/' makes some things easier
214 root = root + os.path.sep
215
216 if not self.__is_path_below(file, root):
217 raise OSError(errno.EINVAL, "file '%s' is not below root" % file)
218
219 try:
220 if use_physdir:
221 file = self.__realpath_rel(root, file[(len(root) - 1):], root, loop_cnt, assume_dir)
222 else:
223 file = self.__realpath(file, root, loop_cnt, assume_dir)[0]
224 except OSError, e:
225 if e.errno == errno.ELOOP:
226 # make ELOOP more readable; without catching it, there will
227 # be printed a backtrace with 100s of OSError exceptions
228 # else
229 raise OSError(errno.ELOOP,
230 "too much recursions while resolving '%s'; loop in '%s'" %
231 (file, e.strerror))
232
233 raise
234
235 return file