From owner-freebsd-emulation Mon Apr 26 8:58:41 1999 Delivered-To: freebsd-emulation@freebsd.org Received: from dan.emsphone.com (dan.emsphone.com [199.67.51.101]) by hub.freebsd.org (Postfix) with ESMTP id 4F6811520A for ; Mon, 26 Apr 1999 08:58:36 -0700 (PDT) (envelope-from dan@dan.emsphone.com) Received: (from dan@localhost) by dan.emsphone.com (8.9.3/8.9.3) id KAA44799; Mon, 26 Apr 1999 10:58:27 -0500 (CDT) (envelope-from dan) Date: Mon, 26 Apr 1999 10:58:26 -0500 From: Dan Nelson To: Mark Newton Cc: freebsd-svr4@atdot.dotat.org, freebsd-emulation@FreeBSD.ORG Subject: Re: [freebsd-svr4] Anyone working on SCO ELF binary emulation? Message-ID: <19990426105826.A44673@dan.emsphone.com> References: <19990424231519.A25153@dan.emsphone.com> <199904260955.TAA21511@atdot.dotat.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.95.3i In-Reply-To: <199904260955.TAA21511@atdot.dotat.org>; from "Mark Newton" on Mon Apr 26 19:25:29 GMT 1999 X-OS: FreeBSD 4.0-CURRENT Sender: owner-freebsd-emulation@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org In the last episode (Apr 26), Mark Newton said: > Dan Nelson wrote: > > I have the duty of making the latest MicroFocus COBOL compiler run > > on FreeBSD, and was wondering if anyone has had any luck getting > > SCO/ELF binaries to run under emulation on -current. > > Yup, in certain limited cases. It still needs work, though. > Interested in helping out? :-) That's why I'm here :) I've got 5.0.2 and 5.0.4 installed on separate boxes, so I have lots of binaries to test with. > Hang on, are we talking about xenix or SCO here? Your ktrace output > includes svr4_sys_* routines (which come from the ELF svr4 emulator), > but you're talking about ibcs2 stuff above. Right. Even the ELF svr4 binaries call syscall #40, which is the SCO xenix syscall. They like calling xenix_eaccess. > > SCO's documentation lists it as SI86GETFEATURES, but of course > > doesn't say what the arguments are, or what I should be returning. > > My guess is that my failure to return valid info for this syscall > > is what's making 'pwd' fail. > > Doubtful. FreeBSD's emulation should work in total ignorance of what > sysarch() returns (although sysarch() is emulated, so if that isn't > the case then values can be fudged which stop executables from > crashing). Even though sysarch() is emulated, the SI86GETFEATURES subfunction is not, so I was wondering if we should be returning an error message, or fabricating some return value. Is returning 0 a valid response, from the SCO app's point of view? Does Linux do svr4 emulation? -Dan Nelson dnelson@emsphone.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-emulation" in the body of the message