From owner-cvs-all Thu Aug 23 12:35:30 2001 Delivered-To: cvs-all@freebsd.org Received: from freebsd.dk (fw-rl0.freebsd.dk [212.242.86.114]) by hub.freebsd.org (Postfix) with ESMTP id 2C2FF37B40D; Thu, 23 Aug 2001 12:35:06 -0700 (PDT) (envelope-from sos@freebsd.dk) Received: (from sos@localhost) by freebsd.dk (8.11.3/8.11.3) id f7NJZL385368; Thu, 23 Aug 2001 21:35:21 +0200 (CEST) (envelope-from sos) From: Søren Schmidt Message-Id: <200108231935.f7NJZL385368@freebsd.dk> Subject: Re: cvs commit: src/sys/nfs nfs_lock.c In-Reply-To: <20010823223936.A571@nagual.pp.ru> "from Andrey A. Chernov at Aug 23, 2001 10:39:37 pm" To: "Andrey A. Chernov" Date: Thu, 23 Aug 2001 21:35:21 +0200 (CEST) Cc: cvs-committers@FreeBSD.org, cvs-all@FreeBSD.org Reply-To: sos@freebsd.dk X-Mailer: ELM [version 2.4ME+ PL88 (25)] MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Content-Type: text/plain; charset=ISO-8859-1 Sender: owner-cvs-all@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG It seems Andrey A. Chernov wrote: > > Excuse me asking, but could we have you do all this testing > > and bugfixing etc BEFORE you start to commit things ?? > > We have now had a dozen or so "oops forgot this forgot that" > > commits all day long, its getting tiresome... > > Only two 'oops' today. All other for implementing just different parts we > don't even have before (EOVERFLOW, l_len<0), cosmetique, spelling, etc. I > prefer small portions over mega-commits with several features at once. > My local tests works, but it is difficult to test full implications of > overflow handling, some fixes come from the people feedback (which I don't > have when post CFR into arch@) It just gets tiresome to try to debug -current, when one has to follow all these little commits, as they broke other parts of the system that influenced where I was chasing bugs, thats why I would have prefered it being tested maybe as patches first by a larger audience, before committing away in a already troublesome -current. Anyhow, its a matter of taste I guess, I'd prefer one commit saying POSIX compliance added or whatever, let it rest with that.... > > Another thing is that this new POSIX behavior breaks some > > apps, take mpg123 as an example, it doesn't work anymore.... > > It means that app must be fixed. POSIX is POSIX. Unless there is mistake > from my side. Check if your vfs_syscalls.c is v1.201 at least - there was > a bug in the code I blindly copy from bde suggestion which breaks lseek > SEEK_END badly. OK, that solved the mpg123 case, thanks! -Søren To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message