Date: Mon, 13 May 2002 15:27:36 -0700 From: Marcel Moolenaar <marcel@cup.hp.com> To: Alexander Kabaev <ak03@gte.com> Cc: freebsd-emulation@FreeBSD.ORG Subject: Re: lahey linux fortran compiler & programs work, but not debugger Message-ID: <20020513222736.GB78999@gauss.cup.hp.com> In-Reply-To: <20020513175727.4c7c920c.ak03@gte.com> References: <200205101513.g4AFDwX1060608@fac13.ds.psu.edu> <20020512234440.GA41169@dhcp01.pn.xcllnt.net> <20020513095734.0a409b1b.ak03@gte.com> <20020513193007.GB42806@gauss.cup.hp.com> <20020513175727.4c7c920c.ak03@gte.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, May 13, 2002 at 05:57:27PM -0400, Alexander Kabaev wrote: > 1. Most obvious shortcoming of the code is that it is i386-specific. I > intended to split the patch in MI/MD parts similar to what native ptrace > implementation is doing, but I haven't got a time for that yet. Not to > mention the fact that I simply have no access to non-i386 FreeBSD boxes > to test my changes. > > 2. There were some locking changes in -CURRENT and the code needs to be > verified to follow the new rules. > > 3. PT_SYSCALL emulation is missing. It would be quite useful to use > Linux's own strace utility to trace Linux binaries. I have semi-working > version of that on my disk somewhere already. > > 4. The future KSE work will require serious rewrite of the ptrace(2) > implementation anyway. Is any of this a blocker? Bullet 1 is less then ideal, but I don't think that not having Alpha support is good reason to block i386 support. In fact, having i386 support will more quickly result in Alpha support than waiting for someone to add Alpha support before we have it at all. Bullet 2 shouldn't take months. In fact, Most of the time we need to spend now is because we let it rot for 3 months. Bullet 3 is opportunistic. Most of the Linuxulator code has been written because there was a need. As such, when there isn't a need, support normally isn't added. As such, missing features are not a blocker to get the working parts in. We also don't create Linux compatible coredumps for example. Bullet 4 is irrelevant. Don't worry about what might happen in the future if it's not more substantial than a rough indication or idea. Anything that hasn't gota detailed schedule is very likely not going to happen at all or anytime soon. Note also that most of this is specific to -current and don't really relate to -stable at all. -- Marcel Moolenaar USPA: A-39004 marcel@xcllnt.net To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-emulation" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20020513222736.GB78999>