From sacadmin Fri May 5 14:18:09 2006 Received: from jurassic.eng.sun.com (jurassic.SFBay.Sun.COM [129.146.68.130]) by sac.sfbay.sun.com (8.12.9+Sun/8.12.9) with ESMTP id k45LI9IQ012121 for ; Fri, 5 May 2006 14:18:09 -0700 (PDT) Received: from [192.9.61.2] (punchin-rie.SFBay.Sun.COM [192.9.61.2]) by jurassic.eng.sun.com (8.13.6+Sun/8.13.6) with ESMTP id k45LI8XQ765143; Fri, 5 May 2006 14:18:09 -0700 (PDT) Message-ID: <445BC03C.2010806@Sun.com> Date: Fri, 05 May 2006 14:14:36 -0700 From: Rod Evans Reply-To: Rod.Evans@Sun.com Organization: Sun Microsystems Inc. User-Agent: Mozilla/5.0 (X11; U; SunOS i86pc; en-US; rv:1.7.12) Gecko/20050925 X-Accept-Language: en-us, en MIME-Version: 1.0 To: psarc@sac.eng.sun.com CC: Ali Bahrami Subject: 2006/301: elf_M32.h removal Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Status: RO Content-Length: 2098 I'm sponsoring the following case for Ali Bahrami. This case qualifies for Architectural self-review, but we wish to record the following information. ----------------------------------------------------------------------- This is a request to remove /usr/include/sys/elf_M32.h from Solaris. I recently had to update dump(1) to have the full set of symbolic names for AMD64 relocation types. While in that code, I discovered the existence of the elf_M32.h header file. This file contains the relocation type codes for the AT&T WE 32100, a CPU that dates from the early 1980's, and which was never productized (according to Wikipedia: http://en.wikipedia.org/wiki/AT&T_Computer_Systems). My best guess is that this file came to us with the original AT&T SVR4 code that was the basis for Solaris 2.x, but that Solaris has never made use of the information in elf_M32.h for a productive purpose. This file is currently accessed within the ON code base by: [] /usr/ccs/bin/dump [] /bin/file [] /usr/include/sys/elf.h I am in the process of putting back a fix to a CR that as a side effect, eliminates the use of this header from dump(1) and file(1). This leaves only elf.h. elf.h includes elf_M32.h (along with all the other platform specific elf_*.h files). As an experiment, I removed elf_M32.h from /usr/include/sys, and modified elf.h not to include it in a pair of local workspaces. I then did a complete nightly builds of the ON consolidation for both sparc and X86 in these workspaces. I found that Solaris makes no additional use of elf_M32.h beyond those listed above, and that it would be a simple matter to remove this file. Normally, I would hesitate to remove a header file with public exposure, even if it was unused by us. In this case however, the extreme age of the CPU it describes, coupled with the fact that machines using that CPU were never widely available, leads me to believe that this file is nothing more than cruft. I am therefore making this request to remove it from Solaris. ----------------------------------------------------------------------- -- Rod From sacadmin Fri May 5 14:42:49 2006 Received: from jurassic.eng.sun.com (jurassic.SFBay.Sun.COM [129.146.68.36]) by sac.sfbay.sun.com (8.12.9+Sun/8.12.9) with ESMTP id k45LgnIQ013440 for ; Fri, 5 May 2006 14:42:49 -0700 (PDT) Received: from hawaiian-sun (vpn-129-150-12-95.SFBay.Sun.COM [129.150.12.95]) by jurassic.eng.sun.com (8.13.6+Sun/8.13.6) with SMTP id k45LgcGe786237; Fri, 5 May 2006 14:42:48 -0700 (PDT) Message-Id: <200605052142.k45LgcGe786237@jurassic.eng.sun.com> Date: Fri, 5 May 2006 11:41:12 -1000 (HST) From: Joseph Kowalski Reply-To: Joseph Kowalski Subject: Re: 2006/301: elf_M32.h removal To: psarc@sac.eng.sun.com, Rod.Evans@sun.com Cc: ali.bahrami@sun.com MIME-Version: 1.0 Content-Type: TEXT/plain; charset=us-ascii Content-MD5: XRHlMBJosrA8NMKI6pAE5Q== X-Mailer: dtmail 1.3.0 @(#)CDE Version 1.7_20 SunOS 5.11 i86pc i386 Status: RO Content-Length: 864 No problem with the proposal, but ... > This file contains the relocation type codes for the AT&T WE 32100, > a CPU that dates from the early 1980's, and which was never productized > (according to Wikipedia: http://en.wikipedia.org/wiki/AT&T_Computer_Systems). Wikipedia is wrong. Since anybody can post there, it often is. I'm very proud that my 14 year old daughter pointed this out to me. It was productized. I even participated in a design/prototype that used it. It was just never successful. (Nice I/O controller chip that went with this.) I'm hazy on this, but I thought the low-end "3B" machines (as in 3B2) were based on this. Even if it wasn't the chip, I'm pretty sure it was a compatible instruction set. If this is correct, we even had some in Sun. All this said, "fry the sucker!". It probably should have been removed in Zeus. - jek3 From sacadmin Fri May 5 14:53:49 2006 Received: from sentry.SFBay.Sun.COM (sentry.SFBay.Sun.COM [129.146.157.28]) by sac.sfbay.sun.com (8.12.9+Sun/8.12.9) with ESMTP id k45LrnIQ013982 for ; Fri, 5 May 2006 14:53:49 -0700 (PDT) Received: from [129.146.11.196] (sr1-umpk-16.SFBay.Sun.COM [129.146.11.196]) by sentry.SFBay.Sun.COM (8.13.3+Sun/8.13.3) with ESMTP id k45LpUoI019658; Fri, 5 May 2006 14:51:30 -0700 (PDT) Message-ID: <445BC982.6020908@Sun.COM> Date: Fri, 05 May 2006 14:54:10 -0700 From: Mike Ditto User-Agent: Mozilla/5.0 (X11; U; SunOS sun4u; en-US; rv:1.7) Gecko/20050530 X-Accept-Language: en-us, en MIME-Version: 1.0 To: Rod.Evans@Sun.COM CC: psarc@sac.eng.sun.com, Ali Bahrami Subject: Re: 2006/301: elf_M32.h removal References: <445BC03C.2010806@Sun.com> In-Reply-To: <445BC03C.2010806@Sun.com> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Status: RO Content-Length: 641 > This file contains the relocation type codes for the AT&T WE 32100, > a CPU that dates from the early 1980's, and which was never productized Not quite. It's used for the whole 32000 family of CPUs which are (were) part of the AT&T reference platform for SVR4, the 3B2. Of course, the only relevance to Solaris even back then is that it's useful to have the ELF tools on all platforms be able to dump objects for other contemporary platforms... and nobody uses 3B2s or any other 32000 system any more. (I actually have a few 3B2/400s in my antiques collection but they haven't conducted current in about 10 years.) -=] Mike [=- From sacadmin Fri May 5 15:28:59 2006 Received: from engmail2sun.Eng.Sun.COM (engmail2sun.SFBay.Sun.COM [129.144.134.19]) by sac.sfbay.sun.com (8.12.9+Sun/8.12.9) with ESMTP id k45MSwIQ018845 for ; Fri, 5 May 2006 15:28:59 -0700 (PDT) Received: from brmea-mail-4.sun.com (brmea-mail-4.Sun.COM [192.18.98.36]) by engmail2sun.Eng.Sun.COM (8.12.10+Sun/8.12.10/ENSMAIL,v2.2) with ESMTP id k45MSwxE001733 for ; Fri, 5 May 2006 15:28:58 -0700 (PDT) Received: from fe-amer-03.sun.com ([192.18.108.177]) by brmea-mail-4.sun.com (8.12.10/8.12.9) with ESMTP id k45MSuIH003849 for ; Fri, 5 May 2006 16:28:58 -0600 (MDT) Received: from conversion-daemon.mail-amer.sun.com by mail-amer.sun.com (Sun Java System Messaging Server 6.2-4.02 (built Sep 9 2005)) id <0IYT00601CF22S00@mail-amer.sun.com> (original mail from Ali.Bahrami@Sun.COM) for psarc@sac.eng.sun.com; Fri, 05 May 2006 16:28:56 -0600 (MDT) Received: from [198.182.198.2] ([199.45.162.234]) by mail-amer.sun.com (Sun Java System Messaging Server 6.2-4.02 (built Sep 9 2005)) with ESMTPSA id <0IYT00N45D472WC0@mail-amer.sun.com>; Fri, 05 May 2006 16:28:55 -0600 (MDT) Date: Fri, 05 May 2006 16:28:47 -0600 From: Ali Bahrami Subject: Re: 2006/301: elf_M32.h removal In-reply-to: <200605052142.k45LgcGe786237@jurassic.eng.sun.com> Sender: Ali.Bahrami@Sun.COM To: Joseph Kowalski Cc: psarc@sac.eng.sun.com, Rod.Evans@Sun.COM Message-id: <445BD19F.2030908@sun.com> MIME-version: 1.0 Content-type: text/plain; format=flowed; charset=ISO-8859-1 Content-transfer-encoding: 7BIT References: <200605052142.k45LgcGe786237@jurassic.eng.sun.com> User-Agent: Thunderbird 1.5.0.2 (Macintosh/20060308) Status: RO Content-Length: 931 >> This file contains the relocation type codes for the AT&T WE 32100, >> a CPU that dates from the early 1980's, and which was never productized >> (according to Wikipedia: http://en.wikipedia.org/wiki/AT&T_Computer_Systems). Well, I should have known better than to use Wikipedia as a main reference, and especially in service of a minor point. I really only wanted to say that we don't need it now. Let me explain how I got there: I found the file, and saw that it was for the EM_M32 elf machine type (in /usr/include/sys/elf.h). The comment in elf.h says that EM_M32 is for "AT&T WE 32100". I remember 3B2s, but never used them (went from vaxes to suns), so I went to Google and then to wikipedia. And, it sounded reasonable ... :-) I should have realized that those relocations were probably for the whole family of AT&T systems from that era. I was misled by the comment referencing the 32100 specifically. - Ali From sacadmin Fri May 5 15:45:09 2006 Received: from sentry.SFBay.Sun.COM (sentry.SFBay.Sun.COM [129.146.157.28]) by sac.sfbay.sun.com (8.12.9+Sun/8.12.9) with ESMTP id k45Mj8IQ020445 for ; Fri, 5 May 2006 15:45:08 -0700 (PDT) Received: from [129.146.11.196] (sr1-umpk-16.SFBay.Sun.COM [129.146.11.196]) by sentry.SFBay.Sun.COM (8.13.3+Sun/8.13.3) with ESMTP id k45Mgm1R019797; Fri, 5 May 2006 15:42:49 -0700 (PDT) Message-ID: <445BD588.8000900@Sun.COM> Date: Fri, 05 May 2006 15:45:28 -0700 From: Mike Ditto User-Agent: Mozilla/5.0 (X11; U; SunOS sun4u; en-US; rv:1.7) Gecko/20050530 X-Accept-Language: en-us, en MIME-Version: 1.0 To: Ali Bahrami CC: Joseph Kowalski , psarc@sac.eng.sun.com, Rod.Evans@Sun.COM Subject: Re: 2006/301: elf_M32.h removal References: <200605052142.k45LgcGe786237@jurassic.eng.sun.com> <445BD19F.2030908@sun.com> In-Reply-To: <445BD19F.2030908@sun.com> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Status: RO Content-Length: 470 Ali Bahrami wrote On 05/05/06 15:28,: > I should have realized that those relocations were probably for the > whole family of AT&T systems from that era. I was misled by the comment > referencing the 32100 specifically. Even the 32100 specifically had substantial relevance at one time (it was in the 3B2/400 among others). The Wikipedia article seems to be saying that the two particular portable computers mentioned weren't ever marketed, not that the 32100 wasn't. From sacadmin Mon May 8 05:16:39 2006 Received: from eastmail2bur.East.Sun.COM (eastmail2bur.East.Sun.COM [129.148.13.40]) by sac.sfbay.sun.com (8.12.9+Sun/8.12.9) with ESMTP id k48CGcIQ000510 for ; Mon, 8 May 2006 05:16:38 -0700 (PDT) Received: from [129.148.226.12] (sr1-unsh01-02.East.Sun.COM [129.148.226.12]) by eastmail2bur.East.Sun.COM (8.12.10+Sun/8.12.10/ENSMAIL,v2.2) with ESMTP id k48CGZWa018798; Mon, 8 May 2006 08:16:36 -0400 (EDT) Message-ID: <445F36A3.5090609@sun.com> Date: Mon, 08 May 2006 08:16:35 -0400 From: Brian Utterback User-Agent: Thunderbird 2.0a1 (X11/20060505) MIME-Version: 1.0 To: Joseph Kowalski CC: psarc@sac.eng.sun.com, Rod.Evans@sun.com, Ali.Bahrami@sun.com Subject: Re: 2006/301: elf_M32.h removal References: <200605052142.k45LgcGe786237@jurassic.eng.sun.com> In-Reply-To: <200605052142.k45LgcGe786237@jurassic.eng.sun.com> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Status: RO Content-Length: 1529 Joseph Kowalski wrote: > No problem with the proposal, but ... > >> This file contains the relocation type codes for the AT&T WE 32100, >> a CPU that dates from the early 1980's, and which was never productized >> (according to Wikipedia: http://en.wikipedia.org/wiki/AT&T_Computer_Systems). > > Wikipedia is wrong. Since anybody can post there, it often is. I'm very > proud that my 14 year old daughter pointed this out to me. > > It was productized. I even participated in a design/prototype that used it. > It was just never successful. (Nice I/O controller chip that went with this.) > > I'm hazy on this, but I thought the low-end "3B" machines (as in 3B2) were > based on this. Even if it wasn't the chip, I'm pretty sure it was a > compatible instruction set. If this is correct, we even had some in Sun. Well, to be fair, Wikipedia is not wrong on this point. Read the linked page again; it does not say that the WE32100 was not productized, it says that the "Companion" and "Alexander" (which were based on the WE32100) were never marketed. It says nothing about the marketing and or productization of the WE32100 itself. And as you noted, the WE32100 was the basis of the low end 3Bx systems so it certainly was both productized and marketed. -- blu Rose are #FF0000, Violets are #0000FF. All my base are belong to you. ---------------------------------------------------------------------- Brian Utterback - OP/N1 RPE, Sun Microsystems, Inc. Ph:877-259-7345, Em:brian.utterback-at-ess-you-enn-dot-kom