X-Rcpt-Trace: gkrellm at lists.netservicesgroup.com
Received-SPF: pass (Cache: Last token {mx} (res=PASS)) client-ip=52.63.116.147; envelope-from=<a.nielsen@shikadi.net>; x-ip-name=sphereful.sorra.shikadi.net;
X-Received: from sphereful.sorra.shikadi.net (unverified [52.63.116.147])
by mailproc.sbbsnet.net (Network-Services-Group-Email) with ESMTP (TLS) id 1265283-1928206
for <gkrellm@lists.netservicesgroup.com>; Sun, 24 May 2020 00:39:33 -0400
X-Return-Path: a.nielsen
DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=shikadi.net
; s=since20200425; h=MIME-Version:Message-ID:Subject:To:From:Date:Sender:
Reply-To:Cc:Content-ID:Content-Description:Resent-Date:Resent-From:
Resent-Sender:Resent-To:Resent-Cc:Resent-Message-ID:In-Reply-To:References;
bh=PHxIPnuJPyn4OpQ+8xm9uk5iZJ1ALFS6+a/tsVaUCIk=; b=fVjmDru8Qw7Tp374LGnj6EDp5
0Pw+hdzv7U1AY/Bs42CfhCzAroxcf/fgVj2FFhYtjyxQPJ+hIIzZ8rFT7goCoAIISecAjsgTUE42T
8W9/NImlaCHMYt9uJ3pthFPd2T2unpgHBsOkhsDDYNHuuZfCo+PpPCkRv/ywo8vpasGGra+UnYaS1
RDSpOUM7RPXkE7z+KDcNvoRcPmGo5ti1pFxkyDXOmh2LoxJKM5lz/i0JjfdN3J1qRKnMRRfKU01EV
+mWOhnlmsBc0MbRFYuivcRyEeNDB+9qrtOcPOehAVqV4cyNRcYYTEUSAmV/qpAp4T8GxLeETwM3lI
phWC9sNZA==;
X-Received: by sphereful.sorra.shikadi.net with esmtpsa (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
(Exim 4.89)
(envelope-from <a.nielsen@shikadi.net>)
id 1jciQ5-0002Z0-Vk
for gkrellm@lists.netservicesgroup.com; Sun, 24 May 2020 04:39:30 +0000
Date: Sun, 24 May 2020 14:39:28 +1000
From: Adam Nielsen
To: gkrellm@lists.netservicesgroup.com
Subject: [Gkrellm] Bug: GKrellM misbehaves with fast networks (integer wraparound?)
Message-ID: <20200524143928.083c2d09@vorticon.teln.shikadi.net>
X-Mailer: Claws Mail 3.17.4 (GTK+ 2.24.32; x86_64-pc-linux-gnu)
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
X-Originating-IP: 52.63.116.147
X-Country: code=AU country="Australia" ip=52.63.116.147
X-ORBS-Accept: dnswl_medium
X-Rcpt-To: <gkrellm@lists.netservicesgroup.com>
X-SpamDetect: *****: 5.0 sd=5.0 Close nspam=0 nok=1 0.00 $0.90(X-Phrase:isspam) $0.10(dnswl_medium) 0.80(X-myrbl:unknown) 0.30(X-NotAscii:us-ascii) 0.69(X-Country:australia) 0.67(X-Mash:sameip) $0.37(genuine) $0.37(dkimok) $0.39(X-Verify-Helo:+OK) $0.42(spfpass) $0.45(StandardTLD) $0.49(X-LangGuess:English) Sane 5.0 NotSaned 7.0 5.0
X-NotAscii: charset=us-ascii
X-Mash: sameip
X-LangGuess: English
X-Phrase: IsSpam score=1.00
X-Verify-Helo: +OK sphereful.sorra.shikadi.net
Authentication-Results: mailproc.sbbsnet.net header.from=a.nielsen@shikadi.net; dkim=pass (good signature)
X-Encryption: SSL encrypted
X-MyRbl: Color=Unknown (rbl) Age=0 Spam=0 Notspam=0 Stars=0 Good=2 Friend=0 Surbl=0 Catch=0 r=0 ip=52.63.116.147
X-IP-stats: Incoming Outgoing Last 0, First 0, in=2, out=4, spam=0 Known=true ip=52.63.116.147
List-Unsubscribe: <mailto:gkrellm-leave@lists.netservicesgroup.com?subject=unsubscribe>
X-Mailing-List: gkrellm@lists.netservicesgroup.com
List-ID: <gkrellm@lists.netservicesgroup.com>
Precedence: bulk
Reply-To: gkrellm@lists.netservicesgroup.com
Hi all,
@Bill I wonder whether you have any ideas how to fix a bug I have just
encountered?
I have bonded two 10 gigabit network interfaces together, which gives
me a connection speed of approx two gigabytes per second. Trying to
display this in GKrellM causes some unusual behaviour:
* The $t and $r labels happily show values up to around "1.7 G" but
then they randomly change to be the letter "a". I am guessing this
might be a problem related to integer size, as around 2 billion is
where 32-bit signed integers wrap around to negatives.
* I have set the 'resolution per grid' to be the maximum 500,000,000
with four grids, so that each grid line represents a quarter of the
available throughput. Given that this is the maximum resolution per
grid, I wonder whether it might need increasing for the future.
Anyone lucky enough to have a 40 Gb or 100 Gb interface won't be
able to fully use GKrellM otherwise :)
* As soon as the 'resolution per grid' hits 500M, there are a dozen
grid lines showing instead of only the four I have set. I am
wondering whether 500M * 4 = 2 billion, which is going into a signed
32-bit integer and coming out as a negative number again. Maybe
these calculations have to become unsigned 64-bit values?
What do you think?
Cheers,
Adam.
|