##// END OF EJS Templates
wireproto: add streams to frame-based protocol...
wireproto: add streams to frame-based protocol Previously, the frame-based protocol was just a series of frames, with each frame associated with a request ID. In order to scale the protocol, we'll want to enable the use of compression. While it is possible to enable compression at the socket/pipe level, this has its disadvantages. The big one is it undermines the point of frames being standalone, atomic units that can be read and written: if you add compression above the framing protocol, you are back to having a stream-based protocol as opposed to something frame-based. So in order to preserve frames, compression needs to occur at the frame payload level. Compressing each frame's payload individually will limit compression ratios because the window size of the compressor will be limited by the max frame size, which is 32-64kb as currently defined. It will also add CPU overhead, as it is more efficient for compressors to operate on fewer, larger blocks of data than more, smaller blocks. So compressing each frame independently is out. This means we need to compress each frame's payload as if it is part of a larger stream. The simplest approach is to have 1 stream per connection. This could certainly work. However, it has disadvantages (documented below). We could also have 1 stream per RPC/command invocation. (This is the model HTTP/2 goes with.) This also has disadvantages. The main disadvantage to one global stream is that it has the very real potential to create CPU bottlenecks doing compression. Networks are only getting faster and the performance of single CPU cores has been relatively flat. Newer compression formats like zstandard offer better CPU cycle efficiency than predecessors like zlib. But it still all too common to saturate your CPU with compression overhead long before you saturate the network pipe. The main disadvantage with streams per request is that you can't reap the benefits of the compression context for multiple requests. For example, if you send 1000 RPC requests (or HTTP/2 requests for that matter), the response to each would have its own compression context. The overall size of the raw responses would be larger because compression contexts wouldn't be able to reference data from another request or response. The approach for streams as implemented in this commit is to support N streams per connection and for streams to potentially span requests and responses. As explained by the added internals docs, this facilitates servers and clients delegating independent streams and compression to independent threads / CPU cores. This helps alleviate the CPU bottleneck of compression. This design also allows compression contexts to be reused across requests/responses. This can result in improved compression ratios and less overhead for compressors and decompressors having to build new contexts. Another feature that was defined was the ability for individual frames within a stream to declare whether that individual frame's payload uses the content encoding (read: compression) defined by the stream. The idea here is that some servers may serve data from a combination of caches and dynamic resolution. Data coming from caches may be pre-compressed. We want to facilitate servers being able to essentially stream bytes from caches to the wire with minimal overhead. Being able to mix and match with frames are compressed within a stream enables these types of advanced server functionality. This commit defines the new streams mechanism. Basic code for supporting streams in frames has been added. But that code is seriously lacking and doesn't fully conform to the defined protocol. For example, we don't close any streams. And support for content encoding within streams is not yet implemented. The change was rather invasive and I didn't think it would be reasonable to implement the entire feature in a single commit. For the record, I would have loved to reuse an existing multiplexing protocol to build the new wire protocol on top of. However, I couldn't find a protocol that offers the performance and scaling characteristics that I desired. Namely, it should support multiple compression contexts to facilitate scaling out to multiple CPU cores and compression contexts should be able to live longer than single RPC requests. HTTP/2 *almost* fits the bill. But the semantics of HTTP message exchange state that streams can only live for a single request-response. We /could/ tunnel on top of HTTP/2 streams and frames with HEADER and DATA frames. But there's no guarantee that HTTP/2 libraries and proxies would allow us to use HTTP/2 streams and frames without the HTTP message exchange semantics defined in RFC 7540 Section 8. Other RPC protocols like gRPC tunnel are built on top of HTTP/2 and thus preserve its semantics of stream per RPC invocation. Even QUIC does this. We could attempt to invent a higher-level stream that spans HTTP/2 streams. But this would be violating HTTP/2 because there is no guarantee that HTTP/2 streams are routed to the same server. The best we can do - which is what this protocol does - is shoehorn all request and response data into a single HTTP message and create streams within. At that point, we've defined a Content-Type in HTTP parlance. It just so happens our media type can also work as a standalone, stream-based protocol, without leaning on HTTP or similar protocol. Differential Revision: https://phab.mercurial-scm.org/D2907

File last commit:

r36667:d3b893ec default
r37304:9bfcbe4f default
Show More
pathutil.py
264 lines | 9.5 KiB | text/x-python | PythonLexer
from __future__ import absolute_import
import errno
import os
import posixpath
import stat
from .i18n import _
from . import (
encoding,
error,
pycompat,
util,
)
def _lowerclean(s):
return encoding.hfsignoreclean(s.lower())
class pathauditor(object):
'''ensure that a filesystem path contains no banned components.
the following properties of a path are checked:
- ends with a directory separator
- under top-level .hg
- starts at the root of a windows drive
- contains ".."
More check are also done about the file system states:
- traverses a symlink (e.g. a/symlink_here/b)
- inside a nested repository (a callback can be used to approve
some nested repositories, e.g., subrepositories)
The file system checks are only done when 'realfs' is set to True (the
default). They should be disable then we are auditing path for operation on
stored history.
If 'cached' is set to True, audited paths and sub-directories are cached.
Be careful to not keep the cache of unmanaged directories for long because
audited paths may be replaced with symlinks.
'''
def __init__(self, root, callback=None, realfs=True, cached=False):
self.audited = set()
self.auditeddir = set()
self.root = root
self._realfs = realfs
self._cached = cached
self.callback = callback
if os.path.lexists(root) and not util.fscasesensitive(root):
self.normcase = util.normcase
else:
self.normcase = lambda x: x
def __call__(self, path, mode=None):
'''Check the relative path.
path may contain a pattern (e.g. foodir/**.txt)'''
path = util.localpath(path)
normpath = self.normcase(path)
if normpath in self.audited:
return
# AIX ignores "/" at end of path, others raise EISDIR.
if util.endswithsep(path):
raise error.Abort(_("path ends in directory separator: %s") % path)
parts = util.splitpath(path)
if (os.path.splitdrive(path)[0]
or _lowerclean(parts[0]) in ('.hg', '.hg.', '')
or pycompat.ospardir in parts):
raise error.Abort(_("path contains illegal component: %s") % path)
# Windows shortname aliases
for p in parts:
if "~" in p:
first, last = p.split("~", 1)
if last.isdigit() and first.upper() in ["HG", "HG8B6C"]:
raise error.Abort(_("path contains illegal component: %s")
% path)
if '.hg' in _lowerclean(path):
lparts = [_lowerclean(p.lower()) for p in parts]
for p in '.hg', '.hg.':
if p in lparts[1:]:
pos = lparts.index(p)
base = os.path.join(*parts[:pos])
raise error.Abort(_("path '%s' is inside nested repo %r")
% (path, pycompat.bytestr(base)))
normparts = util.splitpath(normpath)
assert len(parts) == len(normparts)
parts.pop()
normparts.pop()
prefixes = []
# It's important that we check the path parts starting from the root.
# This means we won't accidentally traverse a symlink into some other
# filesystem (which is potentially expensive to access).
for i in range(len(parts)):
prefix = pycompat.ossep.join(parts[:i + 1])
normprefix = pycompat.ossep.join(normparts[:i + 1])
if normprefix in self.auditeddir:
continue
if self._realfs:
self._checkfs(prefix, path)
prefixes.append(normprefix)
if self._cached:
self.audited.add(normpath)
# only add prefixes to the cache after checking everything: we don't
# want to add "foo/bar/baz" before checking if there's a "foo/.hg"
self.auditeddir.update(prefixes)
def _checkfs(self, prefix, path):
"""raise exception if a file system backed check fails"""
curpath = os.path.join(self.root, prefix)
try:
st = os.lstat(curpath)
except OSError as err:
# EINVAL can be raised as invalid path syntax under win32.
# They must be ignored for patterns can be checked too.
if err.errno not in (errno.ENOENT, errno.ENOTDIR, errno.EINVAL):
raise
else:
if stat.S_ISLNK(st.st_mode):
msg = (_('path %r traverses symbolic link %r')
% (pycompat.bytestr(path), pycompat.bytestr(prefix)))
raise error.Abort(msg)
elif (stat.S_ISDIR(st.st_mode) and
os.path.isdir(os.path.join(curpath, '.hg'))):
if not self.callback or not self.callback(curpath):
msg = _("path '%s' is inside nested repo %r")
raise error.Abort(msg % (path, pycompat.bytestr(prefix)))
def check(self, path):
try:
self(path)
return True
except (OSError, error.Abort):
return False
def canonpath(root, cwd, myname, auditor=None):
'''return the canonical path of myname, given cwd and root
>>> def check(root, cwd, myname):
... a = pathauditor(root, realfs=False)
... try:
... return canonpath(root, cwd, myname, a)
... except error.Abort:
... return 'aborted'
>>> def unixonly(root, cwd, myname, expected='aborted'):
... if pycompat.iswindows:
... return expected
... return check(root, cwd, myname)
>>> def winonly(root, cwd, myname, expected='aborted'):
... if not pycompat.iswindows:
... return expected
... return check(root, cwd, myname)
>>> winonly(b'd:\\\\repo', b'c:\\\\dir', b'filename')
'aborted'
>>> winonly(b'c:\\\\repo', b'c:\\\\dir', b'filename')
'aborted'
>>> winonly(b'c:\\\\repo', b'c:\\\\', b'filename')
'aborted'
>>> winonly(b'c:\\\\repo', b'c:\\\\', b'repo\\\\filename',
... b'filename')
'filename'
>>> winonly(b'c:\\\\repo', b'c:\\\\repo', b'filename', b'filename')
'filename'
>>> winonly(b'c:\\\\repo', b'c:\\\\repo\\\\subdir', b'filename',
... b'subdir/filename')
'subdir/filename'
>>> unixonly(b'/repo', b'/dir', b'filename')
'aborted'
>>> unixonly(b'/repo', b'/', b'filename')
'aborted'
>>> unixonly(b'/repo', b'/', b'repo/filename', b'filename')
'filename'
>>> unixonly(b'/repo', b'/repo', b'filename', b'filename')
'filename'
>>> unixonly(b'/repo', b'/repo/subdir', b'filename', b'subdir/filename')
'subdir/filename'
'''
if util.endswithsep(root):
rootsep = root
else:
rootsep = root + pycompat.ossep
name = myname
if not os.path.isabs(name):
name = os.path.join(root, cwd, name)
name = os.path.normpath(name)
if auditor is None:
auditor = pathauditor(root)
if name != rootsep and name.startswith(rootsep):
name = name[len(rootsep):]
auditor(name)
return util.pconvert(name)
elif name == root:
return ''
else:
# Determine whether `name' is in the hierarchy at or beneath `root',
# by iterating name=dirname(name) until that causes no change (can't
# check name == '/', because that doesn't work on windows). The list
# `rel' holds the reversed list of components making up the relative
# file name we want.
rel = []
while True:
try:
s = util.samefile(name, root)
except OSError:
s = False
if s:
if not rel:
# name was actually the same as root (maybe a symlink)
return ''
rel.reverse()
name = os.path.join(*rel)
auditor(name)
return util.pconvert(name)
dirname, basename = util.split(name)
rel.append(basename)
if dirname == name:
break
name = dirname
# A common mistake is to use -R, but specify a file relative to the repo
# instead of cwd. Detect that case, and provide a hint to the user.
hint = None
try:
if cwd != root:
canonpath(root, root, myname, auditor)
relpath = util.pathto(root, cwd, '')
if relpath[-1] == pycompat.ossep:
relpath = relpath[:-1]
hint = (_("consider using '--cwd %s'") % relpath)
except error.Abort:
pass
raise error.Abort(_("%s not under root '%s'") % (myname, root),
hint=hint)
def normasprefix(path):
'''normalize the specified path as path prefix
Returned value can be used safely for "p.startswith(prefix)",
"p[len(prefix):]", and so on.
For efficiency, this expects "path" argument to be already
normalized by "os.path.normpath", "os.path.realpath", and so on.
See also issue3033 for detail about need of this function.
>>> normasprefix(b'/foo/bar').replace(pycompat.ossep, b'/')
'/foo/bar/'
>>> normasprefix(b'/').replace(pycompat.ossep, b'/')
'/'
'''
d, p = os.path.splitdrive(path)
if len(p) != len(pycompat.ossep):
return path + pycompat.ossep
else:
return path
# forward two methods from posixpath that do what we need, but we'd
# rather not let our internals know that we're thinking in posix terms
# - instead we'll let them be oblivious.
join = posixpath.join
dirname = posixpath.dirname