From: "Daniel P. Berrangé" <berrange@redhat.com>
To: Paolo Bonzini <pbonzini@redhat.com>
Cc: qemu-devel@nongnu.org, "Thomas Huth" <thuth@redhat.com>,
"Ani Sinha" <anisinha@redhat.com>,
"Alex Bennée" <alex.bennee@linaro.org>,
"Philippe Mathieu-Daudé" <philmd@linaro.org>,
"Peter Maydell" <peter.maydell@linaro.org>,
"Cédric Le Goater" <clg@redhat.com>
Subject: Re: [PATCH 13/15] tests/functional: rewrite console handling to be bytewise
Date: Tue, 19 Nov 2024 18:54:05 +0000 [thread overview]
Message-ID: <ZzzezVVCYEJV4uVg@redhat.com> (raw)
In-Reply-To: <beb4abeb-299a-4d4a-a253-3e65a41edcc9@redhat.com>
On Tue, Nov 19, 2024 at 06:11:10PM +0100, Paolo Bonzini wrote:
> On 11/19/24 16:05, Daniel P. Berrangé wrote:
> > The console interaction that waits for predicted strings uses
> > readline(), and thus is only capable of waiting for strings
> > that are followed by a newline.
> >
> > This is inconvenient when needing to match on some things,
> > particularly login prompts, or shell prompts, causing tests
> > to use time.sleep(...) instead, which is unreliable.
> >
> > Switch to reading the console 1 byte at a time, comparing
> > against the success/failure messages until we see a match,
> > regardless of whether a newline is encountered.
> >
> > The success/failure comparisons are done with the python bytes
> > type, rather than strings, to avoid the problem of needing to
> > decode partially received multibyte utf8 characters.
> >
> > Heavily inspired by a patch proposed by Cédric, but written
> > again to work in bytes, rather than strings.
> >
> > Co-developed-by: Cédric Le Goater <clg@redhat.com>
> > Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
> > ---
> > tests/functional/qemu_test/cmd.py | 63 +++++++++++++++++++++++--------
> > 1 file changed, 48 insertions(+), 15 deletions(-)
> >
> > diff --git a/tests/functional/qemu_test/cmd.py b/tests/functional/qemu_test/cmd.py
> > index 76a48064cd..91267a087f 100644
> > --- a/tests/functional/qemu_test/cmd.py
> > +++ b/tests/functional/qemu_test/cmd.py
> > @@ -78,15 +78,58 @@ def run_cmd(args):
> > def is_readable_executable_file(path):
> > return os.path.isfile(path) and os.access(path, os.R_OK | os.X_OK)
> > +def _console_readline(test, vm, success, failure):
> > + msg = bytes([])
> > + done = False
> > + while True:
> > + c = vm.console_socket.recv(1)
> > + if c is None:
> > + done = True
> > + test.fail(
> > + f"EOF in console, expected '{success}'")
> > + break
> > + msg += c
> > +
> > + if success is None or success in msg:
>
> As an optimization, you could use msg.endswith(success) and
> msg.endswith(failure), which would avoid the most blatant cases of O(n^2)
> behavior.
>
> More important, I think "if success is None" should not be here, because it
> will exit after one char. Instead...
>
> > + done = True
> > + break
> > + if failure and failure in msg:
> > + done = True
> > + vm.console_socket.close()
> > + test.fail(
> > + f"'{failure}' found in console, expected '{success}'")
> > +
> > + if c == b'\n':
>
> Here you can put
>
> done = success is None
Hmmm, this can only be a problem if "success" is None, and
"failure" is not None, and although the old code would
technically work in that case, I think it is actually an
unknown/invalid usage scenario.
If BOTH "success" and "failure" are None, this method won't
be called at all. It is valid for "failure" to be none, but
I don't think it makes semantic sense for "success" to also
be None, while have "failure" be non-None.
So I'm inclined to say we declare 'success' to be mandatory
and validate that in the caller. eg
assert send_string is not None or success_message is not None
and then remove this "success is None" check from
_console_readline.
>
> Paolo
>
> > + break
> > +
> > + console_logger = logging.getLogger('console')
> > + try:
> > + console_logger.debug(msg.decode().strip())
> > + except:
> > + console_logger.debug(msg)
> > +
> > + return done
> > +
> > def _console_interaction(test, success_message, failure_message,
> > send_string, keep_sending=False, vm=None):
> > assert not keep_sending or send_string
> > if vm is None:
> > vm = test.vm
> > - console = vm.console_file
> > - console_logger = logging.getLogger('console')
> > +
> > test.log.debug(f"Console interaction success:'{success_message}' " +
> > f"failure:'{failure_message}' send:'{send_string}'")
> > +
> > + # We'll process console in bytes, to avoid having to
> > + # deal with unicode decode errors from receiving
> > + # partial utf8 byte sequences
> > + success_message_b = None
> > + if success_message is not None:
> > + success_message_b = success_message.encode()
> > +
> > + failure_message_b = None
> > + if failure_message is not None:
> > + failure_message_b = failure_message.encode()
> > +
> > while True:
> > if send_string:
> > vm.console_socket.sendall(send_string.encode())
> > @@ -99,20 +142,10 @@ def _console_interaction(test, success_message, failure_message,
> > break
> > continue
> > - try:
> > - msg = console.readline().decode().strip()
> > - except UnicodeDecodeError:
> > - msg = None
> > - if not msg:
> > - continue
> > - console_logger.debug(msg)
> > - if success_message is None or success_message in msg:
> > + if _console_readline(test, vm,
> > + success_message_b,
> > + failure_message_b):
> > break
> > - if failure_message and failure_message in msg:
> > - console.close()
> > - fail = 'Failure message found in console: "%s". Expected: "%s"' % \
> > - (failure_message, success_message)
> > - test.fail(fail)
> > def interrupt_interactive_console_until_pattern(test, success_message,
> > failure_message=None,
>
With regards,
Daniel
--
|: https://berrange.com -o- https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org -o- https://fstop138.berrange.com :|
|: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|
next prev parent reply other threads:[~2024-11-19 18:55 UTC|newest]
Thread overview: 36+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-11-19 15:05 [PATCH 00/15] test/functional: improve functional test debugging & fix tuxrun Daniel P. Berrangé
2024-11-19 15:05 ` [PATCH 01/15] tests/functional: fix mips64el test to honour workdir Daniel P. Berrangé
2024-11-19 15:35 ` Alex Bennée
2024-11-19 17:08 ` Philippe Mathieu-Daudé
2024-11-19 15:05 ` [PATCH 02/15] tests/functional: automatically clean up scratch files after tests Daniel P. Berrangé
2024-11-19 16:21 ` Alex Bennée
2024-11-19 17:28 ` Thomas Huth
2024-11-19 15:05 ` [PATCH 03/15] tests/functional: remove "AVOCADO" from env variable name Daniel P. Berrangé
2024-11-19 16:22 ` Alex Bennée
2024-11-19 17:09 ` Philippe Mathieu-Daudé
2024-11-19 15:05 ` [PATCH 04/15] tests/functional: remove todo wrt avocado.utils.wait_for Daniel P. Berrangé
2024-11-19 16:37 ` Alex Bennée
2024-11-19 15:05 ` [PATCH 05/15] tests/functional: remove leftover :avocado: tags Daniel P. Berrangé
2024-11-19 16:37 ` Alex Bennée
2024-11-19 15:05 ` [PATCH 06/15] tests/functional: remove obsolete reference to avocado bug Daniel P. Berrangé
2024-11-19 16:39 ` Alex Bennée
2024-11-19 15:05 ` [PATCH 07/15] tests/functional: remove comments talking about avocado Daniel P. Berrangé
2024-11-19 17:32 ` Thomas Huth
2024-11-19 15:05 ` [PATCH 08/15] tests/functional: honour self.workdir in ACPI bits tests Daniel P. Berrangé
2024-11-19 17:09 ` Alex Bennée
2024-11-19 15:05 ` [PATCH 09/15] tests/functional: put QEMUMachine logs in testcase log directory Daniel P. Berrangé
2024-11-19 17:10 ` Alex Bennée
2024-11-19 15:05 ` [PATCH 10/15] tests/functional: honour requested test VM name in QEMUMachine Daniel P. Berrangé
2024-11-19 17:40 ` Thomas Huth
2024-11-19 15:05 ` [PATCH 11/15] tests/functional: enable debug logging for QEMUMachine Daniel P. Berrangé
2024-11-21 6:52 ` Thomas Huth
2024-11-19 15:05 ` [PATCH 12/15] tests/functional: logs details of console interaction operations Daniel P. Berrangé
2024-11-21 6:58 ` Thomas Huth
2024-11-19 15:05 ` [PATCH 13/15] tests/functional: rewrite console handling to be bytewise Daniel P. Berrangé
2024-11-19 17:11 ` Paolo Bonzini
2024-11-19 18:54 ` Daniel P. Berrangé [this message]
2024-11-19 19:26 ` Paolo Bonzini
2024-11-19 15:05 ` [PATCH 14/15] tests/functional: remove time.sleep usage from tuxrun tests Daniel P. Berrangé
2024-11-19 15:05 ` [PATCH 15/15] tests/functional: add a QMP backdoor for debugging stalled tests Daniel P. Berrangé
2024-11-21 7:01 ` Thomas Huth
2024-11-19 17:59 ` [PATCH 00/15] test/functional: improve functional test debugging & fix tuxrun Cédric Le Goater
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=ZzzezVVCYEJV4uVg@redhat.com \
--to=berrange@redhat.com \
--cc=alex.bennee@linaro.org \
--cc=anisinha@redhat.com \
--cc=clg@redhat.com \
--cc=pbonzini@redhat.com \
--cc=peter.maydell@linaro.org \
--cc=philmd@linaro.org \
--cc=qemu-devel@nongnu.org \
--cc=thuth@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).