From mboxrd@z Thu Jan 1 00:00:00 1970 X-Spam-Checker-Version: SpamAssassin 3.4.4 (2020-01-24) on inbox.vuxu.org X-Spam-Level: X-Spam-Status: No, score=-0.8 required=5.0 tests=DKIM_ADSP_CUSTOM_MED, DKIM_INVALID,DKIM_SIGNED,FREEMAIL_FROM,HTML_MESSAGE,MAILING_LIST_MULTI, T_SCC_BODY_TEXT_LINE autolearn=ham autolearn_force=no version=3.4.4 Received: (qmail 21363 invoked from network); 29 Mar 2022 15:59:47 -0000 Received: from minnie.tuhs.org (45.79.103.53) by inbox.vuxu.org with ESMTPUTF8; 29 Mar 2022 15:59:47 -0000 Received: by minnie.tuhs.org (Postfix, from userid 112) id E5B4F9D6B6; Wed, 30 Mar 2022 01:59:45 +1000 (AEST) Received: from minnie.tuhs.org (localhost [127.0.0.1]) by minnie.tuhs.org (Postfix) with ESMTP id 8E2C99D6AA; Wed, 30 Mar 2022 01:59:10 +1000 (AEST) Authentication-Results: minnie.tuhs.org; dkim=fail reason="signature verification failed" (2048-bit key; unprotected) header.d=gmail.com header.i=@gmail.com header.b="DFLBEYIx"; dkim-atps=neutral Received: by minnie.tuhs.org (Postfix, from userid 112) id 499D19D6AA; Wed, 30 Mar 2022 01:57:58 +1000 (AEST) Received: from mail-oi1-f170.google.com (mail-oi1-f170.google.com [209.85.167.170]) by minnie.tuhs.org (Postfix) with ESMTPS id 569BE9D684 for ; Wed, 30 Mar 2022 01:57:57 +1000 (AEST) Received: by mail-oi1-f170.google.com with SMTP id r8so19484857oib.5 for ; Tue, 29 Mar 2022 08:57:57 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20210112; h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=RcVPvODA4fSxMElTfzA0MGxTzQxnhsrp6WLwOeFPtb4=; b=DFLBEYIx4rM0wXCCYJgPTo1p6XnnrRroX0w3HVAAdv+ttgKV1/uaTU1s7pI7IeqFIl O/1H47nPq183+geow5lunBnp6iiNsD/nXbBsBpGqcAxP1TZa7vfdLH6APiXYiFwod12A h11FQr7HvpkWeWvh2Sjl4enOGqq23ZhJDUmQ9qykdKRBNIIezu5FSZa7ZIEG8JzzfVQx y5QQjr8m0k8OFqGjTV6xxrk+86Y0G1BX5yKJVS1mZnHNqIfiLu9caN2a9psxVDdVGbBX b8SofWsSq4fjpMSplYii9cc8ku7b0gU+UQJLM07FYt3GJF4r9DErGtBuKJY5mfwibOKv sjjQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=RcVPvODA4fSxMElTfzA0MGxTzQxnhsrp6WLwOeFPtb4=; b=eacy0TpsoMOIB4lhjsc2Ed2MCybSXIVXU/QApn/zR/3PGK2PuzyyYtw926YCtAaB8n On7Et4vGjyHEzdcuK+lITlq6oCJzOX9X1wwahHCKTgWOyFZM2GyWIsUr5vmXtySthCUF l9+wQHGVimb2F4b3SrmQd02SoIsvG+4TEWYredCEheIXmIOou1qTg8pzrv2j9bsxhlT6 3Rdj5s6LniWaQ/r4F7KUqiuKgC7YEpcFe21w70VR64eEIOIyTG6FUUTFCOV+2gGU/APt 0JFCtZ/t18+jiEwroCuzWSJ85mKjsLK62xN9bQ28FcEGdYPKcQUGrxGVdufqiNTMY5tb nmSg== X-Gm-Message-State: AOAM532lIJIvFSbtkoA/4IHn9RwZTpv4ACikxSH7as2pQbb+j1LQoyJd kKKH8OOmRNnrLR9XTmDSkfATVcYed/k3kItZo+UThz4BurU= X-Google-Smtp-Source: ABdhPJwGyIIkz7KxSV84lnOd/jKQWp9y5MsvYKQxrSInH6vUz43wgWaXqISX52xidWFqYLnTogtFf/OwScrLiaY87PY= X-Received: by 2002:a05:6808:1992:b0:2da:7ea7:7ef8 with SMTP id bj18-20020a056808199200b002da7ea77ef8mr162128oib.49.1648569476551; Tue, 29 Mar 2022 08:57:56 -0700 (PDT) MIME-Version: 1.0 References: <9316583A-2461-40B9-8B87-15AC4A719198@planet.nl> In-Reply-To: <9316583A-2461-40B9-8B87-15AC4A719198@planet.nl> From: Dan Cross Date: Tue, 29 Mar 2022 11:57:20 -0400 Message-ID: To: Paul Ruizendaal Content-Type: multipart/alternative; boundary="0000000000000998d305db5d7e55" Subject: Re: [TUHS] Paging code in SysV R2 X-BeenThere: tuhs@minnie.tuhs.org X-Mailman-Version: 2.1.26 Precedence: list List-Id: The Unix Heritage Society mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: TUHS main list Errors-To: tuhs-bounces@minnie.tuhs.org Sender: "TUHS" --0000000000000998d305db5d7e55 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable On Tue, Mar 29, 2022 at 11:25 AM Paul Ruizendaal wrote: > No, sorry, my scope of interest is mostly 1975-1985. > > I did read the Mach virtual memory paper from 1988 - from that paper I > gather that the data structures used are totally different from those in > Sys V or BSD. > > There is also the VM implementation that Richard Miller did on SysV r1 in > 1983. Interestingly, his design seems to parallel the choices made by > Reiser a few years before, but it is lighter touch. Both Reiser and Mille= r > refer to Denning and Tenex as prior art. Miller's 1984 Usenix paper about > this project argues that doing approximated LRU from the page table data > results in a process local working set view, which he argued was preferab= le > to the system global working set view generated in the BSD clock algorith= m. > I brought this up on this list back in 2017, but a few years later, Charles Forsyth in the UK did a VM system for SunOS 4 based on the EMAS system. https://www.terzarima.net/doc/taste.pdf I'm not quite sure when this was written, but it cites papers from 1989, so sometime that year or 1990, would be my guess. - Dan C. > On 29 Mar 2022, at 16:05, Clem Cole wrote: > > > > Fascinating - thank you. > > > > Have you figured out that path from here to the SVR4 code base that was > used for the x86 [which I think also went through a few more generations > after the SVR4 release]? > > > > Clem > > > > On Tue, Mar 29, 2022 at 7:22 AM Paul Ruizendaal via TUHS < > tuhs@minnie.tuhs.org> wrote: > > > > I did not have a lot of time to work on documenting the evolution of > paging / virtual memory code in 32V, Sys III and early SysV in the past > months, but I did get some more background information that seems worth > sharing. > > > > My understanding of the virtual memory story at USG is now as follows: > > > > Somewhere in 1981/82 a project plan for Unix 5 / System V was made and > evolving John Reiser=E2=80=99s virtual memory code for 32V-r3 was part of= that > plan. =E2=80=9CEvolving=E2=80=9D in this context meant making it more mai= ntainable and more > hardware independent. John=E2=80=99s code assumed a memory page, a disk b= lock and a > file block all to be the same size, and it needed to be more general. It > was also designed around the VAX MMU and this too needed to be generalise= d. > The person assigned to that job was Bob (Robert) Baron, reporting to Tom > Raleigh. The project involved quite a bit of re-architecting and progress > was slowish. On top of that Bob left for CMU to work on Mach. Tom Raleigh > tried to pick up where Bob had left off, but progress remained slowish. > > > > In parallel, Keith Kelleman and Steve Burroff were working on Unix for > the 3B20 Unix. They did paging code from scratch around the 3B20 MMU (whi= ch > used a more or less =E2=80=98modern=E2=80=99 page table design) and devel= oped their idea > for the =E2=80=9Cregions=E2=80=9D abstraction to support large, non-conti= guous address > spaces. It seems that they built on the main working set ideas/concepts i= n > the Reiser/Baron/Raleigh code base, combined these with their =E2=80=9Cre= gions=E2=80=9D > idea, made it multi-processor capable and made it all work on the 3B20. > Around that time Tom Raleigh seems to have transferred to Bellcore, and t= he > VAX code base got orphaned. > > > > Two young engineers appear to have picked up the work on the VAX code > base: Dean Jagels and Jim McCormick. My understanding is that they > essentially back ported the 3B20 work to the VAX, falling back on the > Reiser/Baron/Raleigh work where necessary. They got it working, and as fa= r > as I can tell, this is what got released in 1984 as part of SysV R2.4 for > the VAX (the oldest surviving source code for this that I could find). > > > > This somewhat tortuous birth may in part explain why Research chose to > use the 4BSD virtual memory code for 8th edition. > > > > > > --0000000000000998d305db5d7e55 Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable
On Tue, Mar 29, 2022 at 11:25 AM Paul Rui= zendaal <pnr@planet.nl> wrote:
No, sorry, my scope of interest is mostly 1975-1985.

I did read the Mach virtual memory paper from 1988 - from that paper I gath= er that the data structures used are totally different from those in Sys V = or BSD.

There is also the VM implementation that Richard Miller did on SysV r1 in 1= 983.=C2=A0 Interestingly, his design seems to parallel the choices made by = Reiser a few years before, but it is lighter touch. Both Reiser and Miller = refer to Denning and Tenex as prior art. Miller's 1984 Usenix paper abo= ut this project argues that doing approximated LRU from the page table data= results in a process local working set view, which he argued was preferabl= e to the system global working set view generated in the BSD clock algorith= m.

I brought this up on this list back = in 2017, but a few years later, Charles Forsyth in the UK did a VM system f= or SunOS 4 based on the EMAS system.=C2=A0https://www.terzarima.net/doc/taste.pdf
I'm not quite sure when this was written, but it cites pap= ers from 1989, so sometime that year or 1990, would be my guess.
=
=C2=A0 =C2=A0 =C2=A0 =C2=A0 - Dan C.

> On 29 Mar 2022, at 16:0= 5, Clem Cole <clemc@c= cc.com> wrote:
>
> Fascinating - thank you.
>
> Have you figured out that path from here to the SVR4 code base that wa= s used for the x86 [which I think also went through a few more generations = after the SVR4 release]?=C2=A0 =C2=A0
>
> Clem
>
> On Tue, Mar 29, 2022 at 7:22 AM Paul Ruizendaal via TUHS <tuhs@minnie.tuhs.org= > wrote:
>
> I did not have a lot of time to work on documenting the evolution of p= aging / virtual memory code in 32V, Sys III and early SysV in the past mont= hs, but I did get some more background information that seems worth sharing= .
>
> My understanding of the virtual memory story at USG is now as follows:=
>
> Somewhere in 1981/82 a project plan for Unix 5 / System V was made and= evolving John Reiser=E2=80=99s virtual memory code for 32V-r3 was part of = that plan. =E2=80=9CEvolving=E2=80=9D in this context meant making it more = maintainable and more hardware independent. John=E2=80=99s code assumed a m= emory page, a disk block and a file block all to be the same size, and it n= eeded to be more general. It was also designed around the VAX MMU and this = too needed to be generalised. The person assigned to that job was Bob (Robe= rt) Baron, reporting to Tom Raleigh. The project involved quite a bit of re= -architecting and progress was slowish. On top of that Bob left for CMU to = work on Mach. Tom Raleigh tried to pick up where Bob had left off, but prog= ress remained slowish.
>
> In parallel, Keith Kelleman and Steve Burroff were working on Unix for= the 3B20 Unix. They did paging code from scratch around the 3B20 MMU (whic= h used a more or less =E2=80=98modern=E2=80=99 page table design) and devel= oped their idea for the =E2=80=9Cregions=E2=80=9D abstraction to support la= rge, non-contiguous address spaces. It seems that they built on the main wo= rking set ideas/concepts in the Reiser/Baron/Raleigh code base, combined th= ese with their =E2=80=9Cregions=E2=80=9D idea, made it multi-processor capa= ble and made it all work on the 3B20. Around that time Tom Raleigh seems to= have transferred to Bellcore, and the VAX code base got orphaned.
>
> Two young engineers appear to have picked up the work on the VAX code = base: Dean Jagels and Jim McCormick. My understanding is that they essentia= lly back ported the 3B20 work to the VAX, falling back on the Reiser/Baron/= Raleigh work where necessary. They got it working, and as far as I can tell= , this is what got released in 1984 as part of SysV R2.4 for the VAX (the o= ldest surviving source code for this that I could find).
>
> This somewhat tortuous birth may in part explain why Research chose to= use the 4BSD virtual memory code for 8th edition.
>
>

--0000000000000998d305db5d7e55--