From lmaury at gmail.com Thu Sep 12 15:01:48 2013 From: lmaury at gmail.com (=?ISO-8859-1?Q?Lo=EFc_Maury?=) Date: Thu, 12 Sep 2013 15:01:48 +0200 Subject: [FreeVMS] Question about FreeVMS qemu image Message-ID: Hello, I am new to the FreeVMS project, and I have tried the Qemu image. But it seem there are some issues ? Here is the last kernel messages in Qemu : %SYS-I-PAGEFLT, pagefault request from $3C at $0000000000800218 %SYS-I-PAGEFLT, pagefault request from $3C at $0000000000007FE8 %SYS-I-PAGEFLT, pagefault request from $3C at $0000000000801DB0 %SYS-I-PAGEFLT, pagefault request from $3C at $00000000008025F8 %SYS-I-PAGEFLT, pagefault request from $3C at $0000000000300000 %SYS-I-PAGEFLT, pagefault request from $3C at $000000000100C668 %RUN-S-STARTED, PAGER.SYS process started %PAGER-I-SWPADDR, setting swapper base address at $0000000007FFE000 %PAGER-I-SWPSIZE, setting swap size to 0 byte %RUN-S-STOPPED, PAGER.SYS process stopped and Qemu is blocked with this message. Is it normal ? I am on Dell Precision M6300 Intel(R) Core(TM)2 Duo CPU T8300 @ 2.40GH qemu-system-x86_64 Thank you Lo?c -------------- next part -------------- An HTML attachment was scrubbed... URL: From joel.bertrand at systella.fr Thu Sep 12 20:24:13 2013 From: joel.bertrand at systella.fr (=?ISO-8859-1?Q?BERTRAND_Jo=EBl?=) Date: Thu, 12 Sep 2013 20:24:13 +0200 Subject: [FreeVMS] Question about FreeVMS qemu image In-Reply-To: References: Message-ID: <523206CD.3060306@systella.fr> Lo?c Maury a ?crit : > Hello, Hello, > I am new to the FreeVMS project, and I have tried the Qemu image. > > But it seem there are some issues ? > > Here is the last kernel messages in Qemu : > > %SYS-I-PAGEFLT, pagefault request from $3C at $0000000000800218 > %SYS-I-PAGEFLT, pagefault request from $3C at $0000000000007FE8 > %SYS-I-PAGEFLT, pagefault request from $3C at $0000000000801DB0 > %SYS-I-PAGEFLT, pagefault request from $3C at $00000000008025F8 > %SYS-I-PAGEFLT, pagefault request from $3C at $0000000000300000 > %SYS-I-PAGEFLT, pagefault request from $3C at $000000000100C668 > %RUN-S-STARTED, PAGER.SYS process started > %PAGER-I-SWPADDR, setting swapper base address at $0000000007FFE000 > %PAGER-I-SWPSIZE, setting swap size to 0 byte > %RUN-S-STOPPED, PAGER.SYS process stopped It's normal. FreeVMS only creates system threads and tasks. It requires a pager to be usable. And I have found a bug in git tree (SLAB hangs) I haven't fixed until now. Regards, JKB From lmaury at gmail.com Thu Sep 12 21:05:39 2013 From: lmaury at gmail.com (=?ISO-8859-1?Q?Lo=EFc_Maury?=) Date: Thu, 12 Sep 2013 21:05:39 +0200 Subject: [FreeVMS] Question about FreeVMS qemu image In-Reply-To: <523206CD.3060306@systella.fr> References: <523206CD.3060306@systella.fr> Message-ID: On Thu, Sep 12, 2013 at 8:24 PM, BERTRAND Jo?l wrote: > Lo?c Maury a ?crit : > >> Hello, >> > > Hello, > > I am new to the FreeVMS project, and I have tried the Qemu image. >> >> But it seem there are some issues ? >> >> Here is the last kernel messages in Qemu : >> >> %SYS-I-PAGEFLT, pagefault request from $3C at $0000000000800218 >> %SYS-I-PAGEFLT, pagefault request from $3C at $0000000000007FE8 >> %SYS-I-PAGEFLT, pagefault request from $3C at $0000000000801DB0 >> %SYS-I-PAGEFLT, pagefault request from $3C at $00000000008025F8 >> %SYS-I-PAGEFLT, pagefault request from $3C at $0000000000300000 >> %SYS-I-PAGEFLT, pagefault request from $3C at $000000000100C668 >> %RUN-S-STARTED, PAGER.SYS process started >> %PAGER-I-SWPADDR, setting swapper base address at $0000000007FFE000 >> %PAGER-I-SWPSIZE, setting swap size to 0 byte >> %RUN-S-STOPPED, PAGER.SYS process stopped >> > > It's normal. FreeVMS only creates system threads and tasks. It > requires a pager to be usable. And I have found a bug in git tree (SLAB > hangs) I haven't fixed until now. > Ok, thank you for your reply Jo?l. > > Regards, > > JKB > ______________________________**_________________ > FreeVMS mailing list > FreeVMS at rayleigh.systella.fr > https://www.systella.fr/cgi-**bin/mailman/listinfo/freevms > -------------- next part -------------- An HTML attachment was scrubbed... URL: From roart at nvg.ntnu.no Wed Sep 25 15:22:25 2013 From: roart at nvg.ntnu.no (Roar =?iso-8859-1?Q?Thron=E6s?=) Date: Wed, 25 Sep 2013 15:22:25 +0200 Subject: [FreeVMS] Question about FreeVMS qemu image In-Reply-To: <523206CD.3060306@systella.fr> References: <523206CD.3060306@systella.fr> Message-ID: <20130925132225.GC17864@sabre-wulf.nvg.ntnu.no> On Thu, Sep 12, 2013 at 08:24:13PM +0200, BERTRAND Jo?l wrote: > >%SYS-I-PAGEFLT, pagefault request from $3C at $000000000100C668 > >%RUN-S-STARTED, PAGER.SYS process started > >%PAGER-I-SWPADDR, setting swapper base address at $0000000007FFE000 > >%PAGER-I-SWPSIZE, setting swap size to 0 byte > >%RUN-S-STOPPED, PAGER.SYS process stopped > > It's normal. FreeVMS only creates system threads and tasks. It > requires a pager to be usable. And I have found a bug in git tree (SLAB > hangs) I haven't fixed until now. For a couple of years now, the available image has been like this. But does it give a bad impression, or give an incentive to fix it? -- -Roar Thron??s From rmctwo at gmail.com Wed Sep 25 15:58:43 2013 From: rmctwo at gmail.com (Renee) Date: Wed, 25 Sep 2013 09:58:43 -0400 Subject: [FreeVMS] Languages Message-ID: <002601ceb9f7$5be24820$13a6d860$@gmail.com> What language are you using? Renee Culver-former VMS Developer -------------- next part -------------- An HTML attachment was scrubbed... URL: From joel.bertrand at systella.fr Wed Sep 25 16:07:52 2013 From: joel.bertrand at systella.fr (=?ISO-8859-1?Q?BERTRAND_Jo=EBl?=) Date: Wed, 25 Sep 2013 16:07:52 +0200 Subject: [FreeVMS] Question about FreeVMS qemu image In-Reply-To: <20130925132225.GC17864@sabre-wulf.nvg.ntnu.no> References: <523206CD.3060306@systella.fr> <20130925132225.GC17864@sabre-wulf.nvg.ntnu.no> Message-ID: <5242EE38.9020609@systella.fr> Roar Thron?s wrote: > On Thu, Sep 12, 2013 at 08:24:13PM +0200, BERTRAND Jo?l wrote: >>> %SYS-I-PAGEFLT, pagefault request from $3C at $000000000100C668 >>> %RUN-S-STARTED, PAGER.SYS process started >>> %PAGER-I-SWPADDR, setting swapper base address at $0000000007FFE000 >>> %PAGER-I-SWPSIZE, setting swap size to 0 byte >>> %RUN-S-STOPPED, PAGER.SYS process stopped >> >> It's normal. FreeVMS only creates system threads and tasks. It >> requires a pager to be usable. And I have found a bug in git tree (SLAB >> hangs) I haven't fixed until now. > > For a couple of years now, the available image has been like this. > But does it give a bad impression, or give an incentive to fix it? Sure. A sleeping project always gives a bad impression. Since HP announcement, I have received a lot of mails from potential new developers, but until now, no one has proposed patches. Rewrite an operating system is a huge work and I will continue if, and only if, some developers will be able to accept responsabilities (code, translation, documentation, web pages...). Best regards, JKB From joel.bertrand at systella.fr Wed Sep 25 16:14:25 2013 From: joel.bertrand at systella.fr (=?ISO-8859-1?Q?BERTRAND_Jo=EBl?=) Date: Wed, 25 Sep 2013 16:14:25 +0200 Subject: [FreeVMS] Languages In-Reply-To: <002601ceb9f7$5be24820$13a6d860$@gmail.com> References: <002601ceb9f7$5be24820$13a6d860$@gmail.com> Message-ID: <5242EFC1.5060604@systella.fr> Renee wrote: > What language are you using? > > Renee Culver-former VMS Developer C (and functionnal C++ for L4 interface) without char *. There is a string type (pointer + length) to avoid buffer overflow. Regards, JKB From rmctwo at gmail.com Wed Sep 25 16:16:01 2013 From: rmctwo at gmail.com (Renee) Date: Wed, 25 Sep 2013 10:16:01 -0400 Subject: [FreeVMS] Question about FreeVMS qemu image In-Reply-To: <5242EE38.9020609@systella.fr> References: <523206CD.3060306@systella.fr> <20130925132225.GC17864@sabre-wulf.nvg.ntnu.no> <5242EE38.9020609@systella.fr> Message-ID: <003201ceb9f9$c3932a00$4ab97e00$@gmail.com> I am in the process of moving. I am a former VMS developer. I wrote to ask what languages you are using. If it's C, VMS developers HATED C. I would like to help/join the effort but I would recommend Bliss. Renee -----Original Message----- From: FreeVMS [mailto:freevms-bounces at rayleigh.systella.fr] On Behalf Of BERTRAND Jo?l Sent: Wednesday, September 25, 2013 10:08 AM To: FreeVMS mailing list Subject: Re: [FreeVMS] Question about FreeVMS qemu image Roar Thron?s wrote: > On Thu, Sep 12, 2013 at 08:24:13PM +0200, BERTRAND Jo?l wrote: >>> %SYS-I-PAGEFLT, pagefault request from $3C at $000000000100C668 >>> %RUN-S-STARTED, PAGER.SYS process started %PAGER-I-SWPADDR, setting >>> swapper base address at $0000000007FFE000 %PAGER-I-SWPSIZE, setting >>> swap size to 0 byte %RUN-S-STOPPED, PAGER.SYS process stopped >> >> It's normal. FreeVMS only creates system threads and tasks. It >> requires a pager to be usable. And I have found a bug in git tree >> (SLAB >> hangs) I haven't fixed until now. > > For a couple of years now, the available image has been like this. > But does it give a bad impression, or give an incentive to fix it? Sure. A sleeping project always gives a bad impression. Since HP announcement, I have received a lot of mails from potential new developers, but until now, no one has proposed patches. Rewrite an operating system is a huge work and I will continue if, and only if, some developers will be able to accept responsabilities (code, translation, documentation, web pages...). Best regards, JKB _______________________________________________ FreeVMS mailing list FreeVMS at rayleigh.systella.fr https://www.systella.fr/cgi-bin/mailman/listinfo/freevms From joel.bertrand at systella.fr Wed Sep 25 16:26:04 2013 From: joel.bertrand at systella.fr (=?ISO-8859-1?Q?BERTRAND_Jo=EBl?=) Date: Wed, 25 Sep 2013 16:26:04 +0200 Subject: [FreeVMS] Question about FreeVMS qemu image In-Reply-To: <003201ceb9f9$c3932a00$4ab97e00$@gmail.com> References: <523206CD.3060306@systella.fr> <20130925132225.GC17864@sabre-wulf.nvg.ntnu.no> <5242EE38.9020609@systella.fr> <003201ceb9f9$c3932a00$4ab97e00$@gmail.com> Message-ID: <5242F27C.4050301@systella.fr> Renee wrote: > I am in the process of moving. I am a former VMS developer. I wrote to ask > what languages you are using. If it's C, VMS developers HATED C. > I would like to help/join the effort but I would recommend Bliss. I think that we have to build a kernel (or a set of servers in L4 paradigm) that is portable. I'm not sure that bliss is good enough to reach this goal. I'm not sure there is somewhere a open sourced bliss compiler that is stable and usable. And I'm not sure that there are bliss developers enough to start this project with bliss. That being said, I know that VMS users hate C. But I don't know why we cannot use a subset of C functionalities to cut of all "hatable" functionalities. Regards, JKB From mveety at gmail.com Wed Sep 25 22:20:09 2013 From: mveety at gmail.com (Matthew Veety) Date: Wed, 25 Sep 2013 16:20:09 -0400 Subject: [FreeVMS] Question about FreeVMS qemu image In-Reply-To: <5242F27C.4050301@systella.fr> References: <523206CD.3060306@systella.fr> <20130925132225.GC17864@sabre-wulf.nvg.ntnu.no> <5242EE38.9020609@systella.fr> <003201ceb9f9$c3932a00$4ab97e00$@gmail.com> <5242F27C.4050301@systella.fr> Message-ID: On Sep 25, 2013, at 10:26, BERTRAND Jo?l wrote: > > Renee wrote: >> I am in the process of moving. I am a former VMS developer. I wrote to ask >> what languages you are using. If it's C, VMS developers HATED C. >> I would like to help/join the effort but I would recommend Bliss. > > I think that we have to build a kernel (or a set of servers in L4 paradigm) that is portable. I'm not sure that bliss is good enough to reach this goal. I'm not sure there is somewhere a open sourced bliss compiler that is stable and usable. And I'm not sure that there are bliss developers enough to start this project with bliss. > > That being said, I know that VMS users hate C. But I don't know why we cannot use a subset of C functionalities to cut of all "hatable" functionalities. > > Regards, > > JKB Bliss doesn't look terribly hard to parse. We could write one. It could output gas or nasm syntax. Compilers aren't a big deal to write. -- Veety From remi.chateauneu at gmail.com Wed Sep 25 23:04:27 2013 From: remi.chateauneu at gmail.com (rc_gmail_imap) Date: Wed, 25 Sep 2013 22:04:27 +0100 Subject: [FreeVMS] Question about FreeVMS qemu image In-Reply-To: <5242F27C.4050301@systella.fr> References: <523206CD.3060306@systella.fr> <20130925132225.GC17864@sabre-wulf.nvg.ntnu.no> <5242EE38.9020609@systella.fr> <003201ceb9f9$c3932a00$4ab97e00$@gmail.com> <5242F27C.4050301@systella.fr> Message-ID: <52434FDB.70407@gmail.com> Le 25.09.2013 15:26, BERTRAND Jo?l a ?crit : > Renee wrote: >> I am in the process of moving. I am a former VMS developer. I wrote >> to ask >> what languages you are using. If it's C, VMS developers HATED C. >> I would like to help/join the effort but I would recommend Bliss. > > I think that we have to build a kernel (or a set of servers in L4 > paradigm) that is portable. I'm not sure that bliss is good enough to > reach this goal. I'm not sure there is somewhere a open sourced bliss > compiler that is stable and usable. And I'm not sure that there are > bliss developers enough to start this project with bliss. > > That being said, I know that VMS users hate C. But I don't know > why we cannot use a subset of C functionalities to cut of all > "hatable" functionalities. > One of the reason of the success of Unix is that it was written in a portable language: C. C and C++ have been used to program all Unices, Windows, Apple OS (With some Objective-C) etc... On the other hand: http://en.wikipedia.org/wiki/BLISS " It was perhaps the best known systems programming language right up until C made its debut a few years later. Since then, C took off and BLISS faded into obscurity." Of course C was hatable in the 70s when VMS started to live. In the mean time not only it had improved a bit, but now the debate is not using C but C++ for this kind of things. Remi -------------- next part -------------- An HTML attachment was scrubbed... URL: From rmctwo at gmail.com Wed Sep 25 23:21:12 2013 From: rmctwo at gmail.com (Renee) Date: Wed, 25 Sep 2013 17:21:12 -0400 Subject: [FreeVMS] Question about FreeVMS qemu image In-Reply-To: References: <523206CD.3060306@systella.fr> <20130925132225.GC17864@sabre-wulf.nvg.ntnu.no> <5242EE38.9020609@systella.fr> <003201ceb9f9$c3932a00$4ab97e00$@gmail.com> <5242F27C.4050301@systella.fr> Message-ID: <005601ceba35$29969310$7cc3b930$@gmail.com> Could we duplicate the front end of the bliss compiler for this? I feel terrible but compilers were not my thing at all. I have written a generic letter to Oracle sales. They did develop a Windows bliss compiler. I am trying to track that down but I admit - fat chance. I've tried before. Renee -----Original Message----- From: FreeVMS [mailto:freevms-bounces at rayleigh.systella.fr] On Behalf Of Matthew Veety Sent: Wednesday, September 25, 2013 4:20 PM To: FreeVMS mailing list Subject: Re: [FreeVMS] Question about FreeVMS qemu image On Sep 25, 2013, at 10:26, BERTRAND Jo?l wrote: > > Renee wrote: >> I am in the process of moving. I am a former VMS developer. I wrote >> to ask what languages you are using. If it's C, VMS developers HATED C. >> I would like to help/join the effort but I would recommend Bliss. > > I think that we have to build a kernel (or a set of servers in L4 paradigm) that is portable. I'm not sure that bliss is good enough to reach this goal. I'm not sure there is somewhere a open sourced bliss compiler that is stable and usable. And I'm not sure that there are bliss developers enough to start this project with bliss. > > That being said, I know that VMS users hate C. But I don't know why we cannot use a subset of C functionalities to cut of all "hatable" functionalities. > > Regards, > > JKB Bliss doesn't look terribly hard to parse. We could write one. It could output gas or nasm syntax. Compilers aren't a big deal to write. -- Veety _______________________________________________ FreeVMS mailing list FreeVMS at rayleigh.systella.fr https://www.systella.fr/cgi-bin/mailman/listinfo/freevms From rmctwo at gmail.com Wed Sep 25 23:40:56 2013 From: rmctwo at gmail.com (Renee) Date: Wed, 25 Sep 2013 17:40:56 -0400 Subject: [FreeVMS] Question about FreeVMS qemu image In-Reply-To: <52434FDB.70407@gmail.com> References: <523206CD.3060306@systella.fr> <20130925132225.GC17864@sabre-wulf.nvg.ntnu.no> <5242EE38.9020609@systella.fr> <003201ceb9f9$c3932a00$4ab97e00$@gmail.com> <5242F27C.4050301@systella.fr> <52434FDB.70407@gmail.com> Message-ID: <005701ceba37$eb5a62e0$c20f28a0$@gmail.com> ?One of the reason of the success of Unix is that it was written in a portable language: C. C and C++ have been used to program all Unices, Windows, Apple OS (With some Objective-C) etc...? I am not a user. I am a developer from Spitbrook. I did host-based Shadowing. C is used because it was free and given away. It was not because of any logic of C over Bliss as languages although Bliss has not been updated in god knows how long. Renee From: FreeVMS [mailto:freevms-bounces at rayleigh.systella.fr] On Behalf Of rc_gmail_imap Sent: Wednesday, September 25, 2013 5:04 PM To: FreeVMS mailing list Subject: Re: [FreeVMS] Question about FreeVMS qemu image Le 25.09.2013 15:26, BERTRAND Jo?l a ?crit : Renee wrote: I am in the process of moving. I am a former VMS developer. I wrote to ask what languages you are using. If it's C, VMS developers HATED C. I would like to help/join the effort but I would recommend Bliss. I think that we have to build a kernel (or a set of servers in L4 paradigm) that is portable. I'm not sure that bliss is good enough to reach this goal. I'm not sure there is somewhere a open sourced bliss compiler that is stable and usable. And I'm not sure that there are bliss developers enough to start this project with bliss. That being said, I know that VMS users hate C. But I don't know why we cannot use a subset of C functionalities to cut of all "hatable" functionalities. One of the reason of the success of Unix is that it was written in a portable language: C. C and C++ have been used to program all Unices, Windows, Apple OS (With some Objective-C) etc... On the other hand: http://en.wikipedia.org/wiki/BLISS " It was perhaps the best known systems programming language right up until C made its debut a few years later. Since then, C took off and BLISS faded into obscurity." Of course C was hatable in the 70s when VMS started to live. In the mean time not only it had improved a bit, but now the debate is not using C but C++ for this kind of things. Remi -------------- next part -------------- An HTML attachment was scrubbed... URL: From rmctwo at gmail.com Wed Sep 25 23:48:27 2013 From: rmctwo at gmail.com (Renee) Date: Wed, 25 Sep 2013 17:48:27 -0400 Subject: [FreeVMS] VMS Message-ID: <005c01ceba38$f810f980$e832ec80$@gmail.com> DEC only wanted Portability to DEC architectures. C ran on all architectures but performance was compromised. Renee -------------- next part -------------- An HTML attachment was scrubbed... URL: From mveety at gmail.com Thu Sep 26 00:42:46 2013 From: mveety at gmail.com (Matthew Veety) Date: Wed, 25 Sep 2013 18:42:46 -0400 Subject: [FreeVMS] VMS In-Reply-To: <005c01ceba38$f810f980$e832ec80$@gmail.com> References: <005c01ceba38$f810f980$e832ec80$@gmail.com> Message-ID: <2B353E87-22C4-4D44-B5CC-5407C44F07AA@gmail.com> On Sep 25, 2013, at 17:48, "Renee" wrote: > > DEC only wanted Portability to DEC architectures. C ran on all architectures but performance was compromised. > > Renee > _______________________________________________ > FreeVMS mailing list > FreeVMS at rayleigh.systella.fr > https://www.systella.fr/cgi-bin/mailman/listinfo/freevms Performance is compromised by the compiler, not the language. -------------- next part -------------- An HTML attachment was scrubbed... URL: From rmctwo at gmail.com Thu Sep 26 01:39:38 2013 From: rmctwo at gmail.com (Renee) Date: Wed, 25 Sep 2013 19:39:38 -0400 Subject: [FreeVMS] VMS In-Reply-To: <2B353E87-22C4-4D44-B5CC-5407C44F07AA@gmail.com> References: <005c01ceba38$f810f980$e832ec80$@gmail.com> <2B353E87-22C4-4D44-B5CC-5407C44F07AA@gmail.com> Message-ID: <006601ceba48$80081cb0$80185610$@gmail.com> I have nothing good to say of C, I realize that C++ is better but it?s still a C derivative and is a derivative of phone books. Renee From: FreeVMS [mailto:freevms-bounces at rayleigh.systella.fr] On Behalf Of Matthew Veety Sent: Wednesday, September 25, 2013 6:43 PM To: FreeVMS mailing list Subject: Re: [FreeVMS] VMS On Sep 25, 2013, at 17:48, "Renee" > wrote: DEC only wanted Portability to DEC architectures. C ran on all architectures but performance was compromised. Renee _______________________________________________ FreeVMS mailing list FreeVMS at rayleigh.systella.fr https://www.systella.fr/cgi-bin/mailman/listinfo/freevms Performance is compromised by the compiler, not the language. -------------- next part -------------- An HTML attachment was scrubbed... URL: From joel.bertrand at systella.fr Thu Sep 26 08:43:00 2013 From: joel.bertrand at systella.fr (=?ISO-8859-1?Q?BERTRAND_Jo=EBl?=) Date: Thu, 26 Sep 2013 08:43:00 +0200 Subject: [FreeVMS] VMS In-Reply-To: <005c01ceba38$f810f980$e832ec80$@gmail.com> References: <005c01ceba38$f810f980$e832ec80$@gmail.com> Message-ID: <5243D774.4040804@systella.fr> Renee a ?crit : > DEC only wanted Portability to DEC architectures. C ran on all > architectures but performance was compromised. Performances ? Security can be compromised, but not performances. And you can easely avoid security issue. JKB From joel.bertrand at systella.fr Thu Sep 26 08:45:48 2013 From: joel.bertrand at systella.fr (=?windows-1252?Q?BERTRAND_Jo=EBl?=) Date: Thu, 26 Sep 2013 08:45:48 +0200 Subject: [FreeVMS] VMS In-Reply-To: <006601ceba48$80081cb0$80185610$@gmail.com> References: <005c01ceba38$f810f980$e832ec80$@gmail.com> <2B353E87-22C4-4D44-B5CC-5407C44F07AA@gmail.com> <006601ceba48$80081cb0$80185610$@gmail.com> Message-ID: <5243D81C.2070908@systella.fr> Renee a ?crit : > I have nothing good to say of C, I realize that C++ is better but it?s > still a C derivative and is a derivative of phone books. Functional C++ is better than C for some reasons. But standard C++ (object oriented) is not a good language to write an operating system. Object abstraction layers cause a lot of unneeded overhead. Regards, JKB From roart at nvg.ntnu.no Thu Sep 26 11:49:05 2013 From: roart at nvg.ntnu.no (Roar =?iso-8859-1?Q?Thron=E6s?=) Date: Thu, 26 Sep 2013 11:49:05 +0200 Subject: [FreeVMS] Question about FreeVMS qemu image In-Reply-To: References: <523206CD.3060306@systella.fr> <20130925132225.GC17864@sabre-wulf.nvg.ntnu.no> <5242EE38.9020609@systella.fr> <003201ceb9f9$c3932a00$4ab97e00$@gmail.com> <5242F27C.4050301@systella.fr> Message-ID: <20130926094905.GD17864@sabre-wulf.nvg.ntnu.no> On Wed, Sep 25, 2013 at 04:20:09PM -0400, Matthew Veety wrote: > > Bliss doesn't look terribly hard to parse. We could write one. It could output gas or nasm syntax. Compilers aren't a big deal to write. I did implement sufficient parts to be enable to compile for 32-bit the CMUIP telnet client. (64-bit was not possible without many changes.) It is included in the source for FreeVMS 0.3.x. (And how come features like these are still unknown? Strange...) But I wouldn't recommend to continue with Bliss. -- -Roar Thron??s From roart at nvg.ntnu.no Thu Sep 26 14:49:02 2013 From: roart at nvg.ntnu.no (Roar =?iso-8859-1?Q?Thron=E6s?=) Date: Thu, 26 Sep 2013 14:49:02 +0200 Subject: [FreeVMS] Question about FreeVMS qemu image In-Reply-To: <5242F27C.4050301@systella.fr> References: <523206CD.3060306@systella.fr> <20130925132225.GC17864@sabre-wulf.nvg.ntnu.no> <5242EE38.9020609@systella.fr> <003201ceb9f9$c3932a00$4ab97e00$@gmail.com> <5242F27C.4050301@systella.fr> Message-ID: <20130926124902.GE17864@sabre-wulf.nvg.ntnu.no> On Wed, Sep 25, 2013 at 04:26:04PM +0200, BERTRAND Jo?l wrote: > Renee wrote: > >I am in the process of moving. I am a former VMS developer. I wrote to ask > >what languages you are using. If it's C, VMS developers HATED C. > >I would like to help/join the effort but I would recommend Bliss. > > I think that we have to build a kernel (or a set of servers in L4 > paradigm) that is portable. I'm not sure that bliss is good enough to > reach this goal. I'm not sure there is somewhere a open sourced bliss > compiler that is stable and usable. And I'm not sure that there are > bliss developers enough to start this project with bliss. First, I state that I do not know L4, so I can't judge it. (And don't have the time for it.) But how easy/difficult will it be to rewrite/extend to our purposes? How about CPU specific features given in VAX/AXP architecture reference manual, how is it possible to implement those? (For my 0.3, I had modified the kernel's interrupt entry point a bit.) For instance, how will AST, IPL etc be handled with L4? How to fit in the QIO and VMS device driver system with L4? And the other hardware/system-specific issues? -- -Roar Thron??s From rmctwo at gmail.com Thu Sep 26 15:21:27 2013 From: rmctwo at gmail.com (Renee) Date: Thu, 26 Sep 2013 09:21:27 -0400 Subject: [FreeVMS] Question about FreeVMS qemu image In-Reply-To: <20130926124902.GE17864@sabre-wulf.nvg.ntnu.no> References: <523206CD.3060306@systella.fr> <20130925132225.GC17864@sabre-wulf.nvg.ntnu.no> <5242EE38.9020609@systella.fr> <003201ceb9f9$c3932a00$4ab97e00$@gmail.com> <5242F27C.4050301@systella.fr> <20130926124902.GE17864@sabre-wulf.nvg.ntnu.no> Message-ID: <008001cebabb$4ed5fae0$ec81f0a0$@gmail.com> I too am ignorant of L4. Renee -----Original Message----- From: FreeVMS [mailto:freevms-bounces at rayleigh.systella.fr] On Behalf Of Roar Thron?s Sent: Thursday, September 26, 2013 8:49 AM To: FreeVMS mailing list Subject: Re: [FreeVMS] Question about FreeVMS qemu image On Wed, Sep 25, 2013 at 04:26:04PM +0200, BERTRAND Jo?l wrote: > Renee wrote: > >I am in the process of moving. I am a former VMS developer. I wrote > >to ask what languages you are using. If it's C, VMS developers HATED C. > >I would like to help/join the effort but I would recommend Bliss. > > I think that we have to build a kernel (or a set of servers in L4 > paradigm) that is portable. I'm not sure that bliss is good enough to > reach this goal. I'm not sure there is somewhere a open sourced bliss > compiler that is stable and usable. And I'm not sure that there are > bliss developers enough to start this project with bliss. First, I state that I do not know L4, so I can't judge it. (And don't have the time for it.) But how easy/difficult will it be to rewrite/extend to our purposes? How about CPU specific features given in VAX/AXP architecture reference manual, how is it possible to implement those? (For my 0.3, I had modified the kernel's interrupt entry point a bit.) For instance, how will AST, IPL etc be handled with L4? How to fit in the QIO and VMS device driver system with L4? And the other hardware/system-specific issues? -- -Roar Thron??s _______________________________________________ FreeVMS mailing list FreeVMS at rayleigh.systella.fr https://www.systella.fr/cgi-bin/mailman/listinfo/freevms From rmctwo at gmail.com Thu Sep 26 15:31:51 2013 From: rmctwo at gmail.com (Renee) Date: Thu, 26 Sep 2013 09:31:51 -0400 Subject: [FreeVMS] Question about FreeVMS qemu image In-Reply-To: <20130926094905.GD17864@sabre-wulf.nvg.ntnu.no> References: <523206CD.3060306@systella.fr> <20130925132225.GC17864@sabre-wulf.nvg.ntnu.no> <5242EE38.9020609@systella.fr> <003201ceb9f9$c3932a00$4ab97e00$@gmail.com> <5242F27C.4050301@systella.fr> <20130926094905.GD17864@sabre-wulf.nvg.ntnu.no> Message-ID: <008101cebabc$c3075a20$49160e60$@gmail.com> " But I wouldn't recommend to continue with Bliss." Why not? Renrr -----Original Message----- From: FreeVMS [mailto:freevms-bounces at rayleigh.systella.fr] On Behalf Of Roar Thron?s Sent: Thursday, September 26, 2013 5:49 AM To: FreeVMS mailing list Subject: Re: [FreeVMS] Question about FreeVMS qemu image On Wed, Sep 25, 2013 at 04:20:09PM -0400, Matthew Veety wrote: > > Bliss doesn't look terribly hard to parse. We could write one. It could output gas or nasm syntax. Compilers aren't a big deal to write. I did implement sufficient parts to be enable to compile for 32-bit the CMUIP telnet client. (64-bit was not possible without many changes.) It is included in the source for FreeVMS 0.3.x. (And how come features like these are still unknown? Strange...) But I wouldn't recommend to continue with Bliss. -- -Roar Thron??s _______________________________________________ FreeVMS mailing list FreeVMS at rayleigh.systella.fr https://www.systella.fr/cgi-bin/mailman/listinfo/freevms From rmctwo at gmail.com Thu Sep 26 15:37:37 2013 From: rmctwo at gmail.com (Renee) Date: Thu, 26 Sep 2013 09:37:37 -0400 Subject: [FreeVMS] Question about FreeVMS qemu image In-Reply-To: <20130926124902.GE17864@sabre-wulf.nvg.ntnu.no> References: <523206CD.3060306@systella.fr> <20130925132225.GC17864@sabre-wulf.nvg.ntnu.no> <5242EE38.9020609@systella.fr> <003201ceb9f9$c3932a00$4ab97e00$@gmail.com> <5242F27C.4050301@systella.fr> <20130926124902.GE17864@sabre-wulf.nvg.ntnu.no> Message-ID: <008201cebabd$90b1f4d0$b215de70$@gmail.com> How about CPU specific features given in VAX/AXP architecture reference manual, how is it possible to implement those? Like what? Movc3 and Movc5? (For my 0.3, I had modified the kernel's interrupt entry point a bit.) For instance, how will AST, IPL etc be handled with L4? AST completion was IPL based at IPL 2 or was that I/O completion? Renee -----Original Message----- From: FreeVMS [mailto:freevms-bounces at rayleigh.systella.fr] On Behalf Of Roar Thron?s Sent: Thursday, September 26, 2013 8:49 AM To: FreeVMS mailing list Subject: Re: [FreeVMS] Question about FreeVMS qemu image On Wed, Sep 25, 2013 at 04:26:04PM +0200, BERTRAND Jo?l wrote: > Renee wrote: > >I am in the process of moving. I am a former VMS developer. I wrote > >to ask what languages you are using. If it's C, VMS developers HATED C. > >I would like to help/join the effort but I would recommend Bliss. > > I think that we have to build a kernel (or a set of servers in L4 > paradigm) that is portable. I'm not sure that bliss is good enough to > reach this goal. I'm not sure there is somewhere a open sourced bliss > compiler that is stable and usable. And I'm not sure that there are > bliss developers enough to start this project with bliss. First, I state that I do not know L4, so I can't judge it. (And don't have the time for it.) But how easy/difficult will it be to rewrite/extend to our purposes? How about CPU specific features given in VAX/AXP architecture reference manual, how is it possible to implement those? (For my 0.3, I had modified the kernel's interrupt entry point a bit.) For instance, how will AST, IPL etc be handled with L4? How to fit in the QIO and VMS device driver system with L4? And the other hardware/system-specific issues? -- -Roar Thron??s _______________________________________________ FreeVMS mailing list FreeVMS at rayleigh.systella.fr https://www.systella.fr/cgi-bin/mailman/listinfo/freevms From joel.bertrand at systella.fr Thu Sep 26 16:08:11 2013 From: joel.bertrand at systella.fr (=?ISO-8859-1?Q?BERTRAND_Jo=EBl?=) Date: Thu, 26 Sep 2013 16:08:11 +0200 Subject: [FreeVMS] Question about FreeVMS qemu image In-Reply-To: <20130926124902.GE17864@sabre-wulf.nvg.ntnu.no> References: <523206CD.3060306@systella.fr> <20130925132225.GC17864@sabre-wulf.nvg.ntnu.no> <5242EE38.9020609@systella.fr> <003201ceb9f9$c3932a00$4ab97e00$@gmail.com> <5242F27C.4050301@systella.fr> <20130926124902.GE17864@sabre-wulf.nvg.ntnu.no> Message-ID: <52443FCB.3000407@systella.fr> Roar Thron?s wrote: > On Wed, Sep 25, 2013 at 04:26:04PM +0200, BERTRAND Jo?l wrote: >> Renee wrote: >>> I am in the process of moving. I am a former VMS developer. I wrote to ask >>> what languages you are using. If it's C, VMS developers HATED C. >>> I would like to help/join the effort but I would recommend Bliss. >> >> I think that we have to build a kernel (or a set of servers in L4 >> paradigm) that is portable. I'm not sure that bliss is good enough to >> reach this goal. I'm not sure there is somewhere a open sourced bliss >> compiler that is stable and usable. And I'm not sure that there are >> bliss developers enough to start this project with bliss. > > First, I state that I do not know L4, so I can't judge it. > (And don't have the time for it.) > > But how easy/difficult will it be to rewrite/extend to our purposes? L4/X2 has to be used without modification. 0.4 should be a 0.3 kernel split in several L4 servers. Main L4 interest is to run several kernel threads in different address spaces. > How about CPU specific features given in VAX/AXP architecture reference > manual, how is it possible to implement those? Yes, it is possible. > (For my 0.3, I had modified the kernel's interrupt entry point a bit.) > For instance, how will AST, IPL etc be handled with L4? > > How to fit in the QIO and VMS device driver system with L4? > > And the other hardware/system-specific issues? L4 does not know any driver. It only handle few syscalls (IPC, threads and very basic memory management). All drivers have to be written as L4 servers. Regards, JKB From roart at nvg.ntnu.no Thu Sep 26 19:27:06 2013 From: roart at nvg.ntnu.no (Roar =?iso-8859-1?Q?Thron=E6s?=) Date: Thu, 26 Sep 2013 19:27:06 +0200 Subject: [FreeVMS] Question about FreeVMS qemu image In-Reply-To: <008201cebabd$90b1f4d0$b215de70$@gmail.com> References: <523206CD.3060306@systella.fr> <20130925132225.GC17864@sabre-wulf.nvg.ntnu.no> <5242EE38.9020609@systella.fr> <003201ceb9f9$c3932a00$4ab97e00$@gmail.com> <5242F27C.4050301@systella.fr> <20130926124902.GE17864@sabre-wulf.nvg.ntnu.no> <008201cebabd$90b1f4d0$b215de70$@gmail.com> Message-ID: <20130926172706.GF17864@sabre-wulf.nvg.ntnu.no> On Thu, Sep 26, 2013 at 09:37:37AM -0400, Renee wrote: > How about CPU specific features given in VAX/AXP architecture reference > manual, how is it possible to implement those? > > Like what? Movc3 and Movc5? AST, IPL, queueing etc. -- -Roar Thron??s From roart at nvg.ntnu.no Thu Sep 26 19:36:18 2013 From: roart at nvg.ntnu.no (Roar =?iso-8859-1?Q?Thron=E6s?=) Date: Thu, 26 Sep 2013 19:36:18 +0200 Subject: [FreeVMS] Question about FreeVMS qemu image In-Reply-To: <008101cebabc$c3075a20$49160e60$@gmail.com> References: <523206CD.3060306@systella.fr> <20130925132225.GC17864@sabre-wulf.nvg.ntnu.no> <5242EE38.9020609@systella.fr> <003201ceb9f9$c3932a00$4ab97e00$@gmail.com> <5242F27C.4050301@systella.fr> <20130926094905.GD17864@sabre-wulf.nvg.ntnu.no> <008101cebabc$c3075a20$49160e60$@gmail.com> Message-ID: <20130926173618.GG17864@sabre-wulf.nvg.ntnu.no> On Thu, Sep 26, 2013 at 09:31:51AM -0400, Renee wrote: > " But I wouldn't recommend to continue with Bliss." > > Why not? It is a bit too low-level, and more CPU-dependent. In the recent OpenVMS development, C took its place (I think). The current Bliss was for GCC suite 4.2.1, the current is at 4.7/4.8, it will need some extra effort to keep up with the GCC releases. (It may take some effort to keep up with the GCC releases anyway, even when relating to C source.) But it is not my decision, whether to use it or not. -- -Roar Thron??s From rmctwo at gmail.com Thu Sep 26 23:18:03 2013 From: rmctwo at gmail.com (Renee) Date: Thu, 26 Sep 2013 17:18:03 -0400 Subject: [FreeVMS] Question about FreeVMS qemu image In-Reply-To: <20130926172706.GF17864@sabre-wulf.nvg.ntnu.no> References: <523206CD.3060306@systella.fr> <20130925132225.GC17864@sabre-wulf.nvg.ntnu.no> <5242EE38.9020609@systella.fr> <003201ceb9f9$c3932a00$4ab97e00$@gmail.com> <5242F27C.4050301@systella.fr> <20130926124902.GE17864@sabre-wulf.nvg.ntnu.no> <008201cebabd$90b1f4d0$b215de70$@gmail.com> <20130926172706.GF17864@sabre-wulf.nvg.ntnu.no> Message-ID: <000f01cebafd$e3355ae0$a9a010a0$@gmail.com> Those are software functions although they do depend on IPL which is a hardware function. Queuing depends on insque/remque which are hardware functions (instructions) but they can be emulated. It is most important that IPL be very carefully thought through. Renee -----Original Message----- From: FreeVMS [mailto:freevms-bounces at rayleigh.systella.fr] On Behalf Of Roar Thron?s Sent: Thursday, September 26, 2013 1:27 PM To: FreeVMS mailing list Subject: Re: [FreeVMS] Question about FreeVMS qemu image On Thu, Sep 26, 2013 at 09:37:37AM -0400, Renee wrote: > How about CPU specific features given in VAX/AXP architecture > reference manual, how is it possible to implement those? > > Like what? Movc3 and Movc5? AST, IPL, queueing etc. -- -Roar Thron??s _______________________________________________ FreeVMS mailing list FreeVMS at rayleigh.systella.fr https://www.systella.fr/cgi-bin/mailman/listinfo/freevms From rmctwo at gmail.com Thu Sep 26 23:28:57 2013 From: rmctwo at gmail.com (Renee) Date: Thu, 26 Sep 2013 17:28:57 -0400 Subject: [FreeVMS] Bliss Message-ID: <001001cebaff$6953f5e0$3bfbe1a0$@gmail.com> "It is a bit too low-level, and more CPU-dependent." I never wrote any high level code except for in the benchmark center in DC. I did not write anything in Bliss there. But I did know it. As far as CPU dependency, there all of my stuff was in Fortran-iV+ I see nothing wrong with revlocking onto Intel. We are at the correct level for Bliss. Renee -------------- next part -------------- An HTML attachment was scrubbed... URL: From roart at nvg.ntnu.no Fri Sep 27 07:44:10 2013 From: roart at nvg.ntnu.no (Roar =?iso-8859-1?Q?Thron=E6s?=) Date: Fri, 27 Sep 2013 07:44:10 +0200 Subject: [FreeVMS] Question about FreeVMS qemu image In-Reply-To: <000f01cebafd$e3355ae0$a9a010a0$@gmail.com> References: <523206CD.3060306@systella.fr> <20130925132225.GC17864@sabre-wulf.nvg.ntnu.no> <5242EE38.9020609@systella.fr> <003201ceb9f9$c3932a00$4ab97e00$@gmail.com> <5242F27C.4050301@systella.fr> <20130926124902.GE17864@sabre-wulf.nvg.ntnu.no> <008201cebabd$90b1f4d0$b215de70$@gmail.com> <20130926172706.GF17864@sabre-wulf.nvg.ntnu.no> <000f01cebafd$e3355ae0$a9a010a0$@gmail.com> Message-ID: <20130927054410.GH17864@sabre-wulf.nvg.ntnu.no> On Thu, Sep 26, 2013 at 05:18:03PM -0400, Renee wrote: > Those are software functions although they do depend on IPL which is a > hardware function. > It is most important that IPL be very carefully thought through. The REI return from exception or interrupt may request an AST, and it will also handle software interrupts. (I put simulation on that together with each Intel interrupt return.) -- -Roar Thron??s