Curt Arnold wrote:
No it is still helpful. I threw together the CVS stuff in a few hours
to keep the conversation going and getting it to work and getting the
API right were the priorities. Deriving log4cxx::basic_logstreambuf
from std::basic_stringbuf was expedient to get things functional, b
I'm sorry, I appear to have wasted my time and yours. When you asked
for construction metrics I had assumed you meant for your
implementation. With your current scheme there would be no way to
avoid the cost of instantiating std::basic_stringbuf, as it's not a
templated parameter.
No it is sti
Curt Arnold wrote:
The benchmark, like all benchmarks, is synthetic. This one assumes
that a log stream is allocated for each logging request which I would
assume is very atypical. I'd expect that most of the cost is in the
construction of std::basic_stringbuf, but would love to see a profile
On my home system (2.4GHz P4 running Gentoo & gcc-3.3) these are the
numbers I got (this is the average time per iteration):
CVS logstream
Logging when disabled: 1600-1800ns
Logging when enabled: 13000-15000ns
My LoggingStream
Logging when disabled: 40-50ns
Logging when enabled: 15000-18000ns
Th
Christopher Smith wrote:
P.S.: I just noticed some performance optimizations to LocationInfo
committed into CVS. I'll rerun and see if they have any impact on
these results.
Looks like it shaved off about 100ns per iteration across the board.
--Chris
Christopher Smith wrote:
Curt Arnold wrote:
On Oct 7, 2004, at 2:51 PM, Christopher Smith wrote:
1) You pay the price of constructing std::basic_stream regardless of
whether logging is turned on.
Do you have any metrics on the cost of that?
No, but I may have some in a week or so. I'll post them
Curt Arnold wrote:
I mentioned this back in our earlier emails
(http://nagoya.apache.org/eyebrowse/ReadMsg?listName=log4cxx-
[EMAIL PROTECTED]&msgNo=272, see point 1), I was surprised when
your submission didn't derive from basic_ostream.
You were thinking inheritence and I was thinking deleg
On Oct 7, 2004, at 4:47 PM, Christopher Smith wrote:
Curt Arnold wrote:
On Oct 7, 2004, at 2:51 PM, Christopher Smith wrote:
1) You pay the price of constructing std::basic_stream regardless of
whether logging is turned on.
Do you have any metrics on the cost of that?
No, but I may have some in a
Curt Arnold wrote:
On Oct 7, 2004, at 2:51 PM, Christopher Smith wrote:
1) You pay the price of constructing std::basic_stream regardless of
whether logging is turned on.
Do you have any metrics on the cost of that?
No, but I may have some in a week or so. I'll post them as soon as I do.
I don't t
On Oct 7, 2004, at 2:51 PM, Christopher Smith wrote:
Curt Arnold wrote:
Some additional comments on the patch are available at the bug
report, LOGCXX-18. The major difference is that this logstream is
derived from std::basic_stream and can be passed to methods expecting
that stream. The inser
Curt Arnold wrote:
Some additional comments on the patch are available at the bug report,
LOGCXX-18. The major difference is that this logstream is derived from
std::basic_stream and can be passed to methods expecting that stream.
The insertion operator will typically be short-circuited when
I've committed my stab at logging streams to the CVS. The stream based
interface is defined in include/log4cxx/stream.h and there is a usage
example in examples/stream.cpp. I've updated the Makefile.am's so the
make based build will work (after an autogen), but I haven't updated
the MSVC proj
Thanks for the submission. It wasn't quite what I had in mind. I'm
trying to flesh my thoughts out right now (well after getting some
sleep) and will reply tomorrow.
On Oct 6, 2004, at 1:31 PM, Christopher Smith wrote:
Curt Arnold wrote:
Thanks, Please submit it as a bug request since that has
Curt Arnold wrote:
Thanks, Please submit it as a bug request since that has a explicit
statement that you are donating the code to ASF and I'll get it in
shortly.
Done. ID is LOGCXX-18.
I fixed forcedLog before submitting it.
--Chris
Thanks, Please submit it as a bug request since that has a explicit
statement that you are donating the code to ASF and I'll get it in
shortly.
Christopher Smith wrote:
7) The code is specifically designed so an optimizer can produce code
that only checks logging levels once per log event.
Ugh. I knew I'd make a mistake. ;-)
This is not quite true. The current code actually requires 2 checks per
log event. This is because on the implemen
16 matches
Mail list logo
