Date: Fri, 19 Feb 2016 10:43:56 +0300 From: Svyatoslav Razmyslov <razmyslov@viva64.com> To: Russell Haley <russ.haley@gmail.com>, Warren Block <wblock@wonkity.com> Cc: Fehmi Noyan ISI <fnoyanisi@yahoo.com>, Andrey Karpov <karpov@viva64.com>, freebsd-doc@FreeBSD.org Subject: Re: Checking FreeBSD kernel with static analyzer Message-ID: <56C6C7BC.2050302@viva64.com> In-Reply-To: <20160218165425.4382803.23287.3091@gmail.com> References: <20160218165425.4382803.23287.3091@gmail.com>
index | next in thread | previous in thread | raw e-mail
We used the same compiler to build the kernel and to preproccess the files, but adding the '-E' flag to the compilation parameters. 18.02.2016 19:54, Russell Haley пишет: > Okay, are those false positives being documented somehow for future reference? If you don't have a bug report nobody knows next time the analyzer is run. Or worse an hack like me is looking at the code! even a code comment might be a good idea about why it's okay? > > Not releasing the bug list immediately was a wise decision IMHO. > > Svyatoslav, Is your pre-processor harness open or closed source? > > Russ > > Sent from my BlackBerry 10 smartphone on the Koodo network. > Original Message > From: Warren Block > Sent: Thursday, February 18, 2016 6:47 AM > To: Svyatoslav Razmyslov > Cc: Russell Haley; Fehmi Noyan ISI; Andrey Karpov; freebsd-doc@FreeBSD.org > Subject: Re: Checking FreeBSD kernel with static analyzer > > On Thu, 18 Feb 2016, Svyatoslav Razmyslov wrote: > >> I have sent the file in CSV format (PVS-Studio-log-freebsd.csv). Does it work >> for you? >> >> 18.02.2016 11:24, Russell Haley пишет: >>> Svyatoslav, >>> >>> Is there an output format for the issue report that could be massaged in >>> python or lua or the like and submitted or imported through bugzilla? CSV, >>> xml, json? > The mailing list stripped the attachment, although the mail direct to me > included it. The review has been of the text file. We can make the CSV > available. > > Automated creation of bug reports from these results might not be too > useful. Many are only suspected problems and turn out to be false > positives. > -- Svyatoslav Razmyslov "Program Verification Systems" (Co Ltd)help
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?56C6C7BC.2050302>