From owner-freebsd-hackers Sun Dec 17 11:59:19 1995 Return-Path: owner-hackers Received: (from root@localhost) by freefall.freebsd.org (8.7.3/8.7.3) id LAA15099 for hackers-outgoing; Sun, 17 Dec 1995 11:59:19 -0800 (PST) Received: from haywire.DIALix.COM (news@haywire.DIALix.COM [192.203.228.65]) by freefall.freebsd.org (8.7.3/8.7.3) with ESMTP id LAA15088 for ; Sun, 17 Dec 1995 11:59:10 -0800 (PST) Received: (from news@localhost) by haywire.DIALix.COM (sendmail) id DAA23336 for freebsd-hackers@freebsd.org; Mon, 18 Dec 1995 03:58:58 +0800 (WST) Received: from GATEWAY by haywire.DIALix.COM with netnews for freebsd-hackers@freebsd.org (problems to: usenet@haywire.dialix.com) To: freebsd-hackers@freebsd.org Date: 18 Dec 1995 03:58:54 +0800 From: peter@haywire.dialix.com (Peter Wemm) Message-ID: <4b1spu$mou$1@haywire.DIALix.COM> Organization: DIALix Services, Perth, Australia. References: <199512171436.OAA13800@server.netcraft.co.uk>, <199512171731.JAA09419@freefall.freebsd.org> Reply-To: peter@jhome.dialix.com Subject: Re: FreeBSD-current-stable ??? Sender: owner-hackers@freebsd.org Precedence: bulk dyson@freefall.freebsd.org (John Dyson) writes: >> Absolutely, the vm headers are a good example. Peter was fixing >> code for a while after this commit because some parts of the tree >> WOULDN'T EVEN COMPILE! Was a make world done before this commit, >> I somehow doubt it given the glaring problems that Peter fixed. >> Missing the odd bug when changing the filesystem to allow 1Tb files >> is one thing and is what -current is for, not doing even a basic >> sanity check to make sure the tree still compiles is a totally >> different case and one we used to be a lot more stern about when >> it happened. >> >> I think the core team has become a little too soft when dealing >> with it's cotributors :-) >> >A make world was not done, and if someone would donate a reasonable >machine to me to let me do so -- it would be very nice. I probably >have the least powerful machine of any major contributor (and have >only one.) Those with expensive high power machines are welcome to >help. (Machines bigger than a 3 yr old 20MB 486/66 :-)). I'm using Julian Elischer's 16MB 386SX-25 box with my own 486/66 motherboard plus some memory in it.. I know the feeling! (hence the reason why I committed the "make -DNOCLEAN world" speedup.. :-) Most of the significant changes that I've been making have been running for a while, some on my employer's SVR4 machines in different forms.. ;-) I try and resist the temptation to make "harmless quick fixes" on the spur of the moment, but I've been a bit less careful lately and have done a few bad ones.. >The kernel did work -- and there was some chaff (a bug in sys_process.c) >-- oh, by the way did the 1Tb changes break things -- or was it the header file >changes/improvements???? All I had to do to get ps working again was to >rebuild libkvm/ps.... The 1Tb kernel stuff has fine as far as I could tell... It's been working flawlessly for me since it went in. The include file cleanup was what caused the most pain that I could see. Unfortunately, the extreme nesting of include files was directly linked to an unnecessary compile slowdown.. It _had_ to be cleaned up. The only problem is that because of the excessive nesting and redundant or inefficient choices that's accumulated over time, many of the user-mode programs have become dependent on this junk. Sure, it was painful while it was up in the air, and a good deal of it was not obvious how to fix for the non-guru types who have been plodding along in -current, but I think it is/was worth it... Currently, as far as I can tell, the main horror stories being reported at the moment appear to be coming from AHC driver users (2740/2840/2940).. Oh! BTW: My system was reliably panicing if I did a 'dd' from a raw partition with a 1MB block size, in vm_bounce_page_free(). This has inexplicably stopped some time over the last week or so!! (Buslogic BT445S VLB card with inability to DMA above 16MB, so bounce buffers are used). -Peter >John >dyson@freebsd.org