Date: Sun, 17 Dec 2000 15:06:50 +0100 From: Szilveszter Adam <sziszi@petra.hos.u-szeged.hu> To: freebsd-stable@FreeBSD.ORG Subject: Re: mozilla and 4.2-RELEASE Message-ID: <20001217150650.B19311@petra.hos.u-szeged.hu> In-Reply-To: <20001217051120.A20812@citusc.usc.edu>; from kris@FreeBSD.ORG on Sun, Dec 17, 2000 at 05:11:20AM -0800 References: <976986477.3a3ba16db730c@webmail.harmonic.co.il> <20001217051120.A20812@citusc.usc.edu>
next in thread | previous in thread | raw e-mail | index | archive | help
Hi! On Sun, Dec 17, 2000 at 05:11:20AM -0800, Kris Kennaway wrote: > Previous versions of mozilla needed to be run as root the first time > so they can create some files in the install directory. I don't know > if this is still the case. Yes, for creation of the registry file. But this is necessary only once. Alternatively, you can just chmod. > Mozilla is also notoriously picky about old versions of files in > ~/.mozilla - try blowing away that directory as well. Maybe, but not likely. I have meant to chime in to this thread earlier, but since I run -CURRENT, my findings may or may not be relevant. I used to regularly build Mozilla from cvs and it mostly worked. The last working Mozilla was sometime end of October, maybe a bit later. (I know for sure that the compile on 13th October worked but I also built it later on and for some time with success.) The only issue after M18 came out was that I could not get PSM to work on FreeBSD whereas it worked on Linux. But this is not a major problem, because the binary versions of PSM for Windows and Linux that are available from iPlanet are not built from the publicly available sources but rather from a tainted version and linked against the RSA BSafe toolkit. PSM did build but did not work. Odd... However, sometime in the middle of November I did build again. The build went through with no problems, but the resulting binary segfaulted on start. At that time there was some speculation going on -current@ wrt bugs after changing the .crtbegin and friends to stock GCC instead of our own, so I assumed it could be just that. Later the issue of linking against -lgcc_r came to light, so I thought this might be related. And after all, this is -CURRENT, so things might break for a while. But the situation has not improved since. The binary builds seemingly just fine, but segfaults on startup. A gdb(1) trace shows that the segfault is in the internal functions of nspr, so it is possibly not a C++ problem. Once the stack trace went as deep as libc_r, however. Since I did not build with debugging symbols, I did not investigate further. Since the Mozilla team has dropped all FreeBSD tinderboxes from their cluster (although they have added OpenBSD 2.5 back) they are probably totally unaware of the problems. -- Regards: Szilveszter ADAM Szeged University Szeged Hungary To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20001217150650.B19311>