From owner-cvs-sys Sun Feb 16 00:23:31 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id AAA19548 for cvs-sys-outgoing; Sun, 16 Feb 1997 00:23:31 -0800 (PST) Received: from spinner.DIALix.COM (spinner.DIALix.COM [192.203.228.67]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id AAA19518; Sun, 16 Feb 1997 00:22:04 -0800 (PST) Received: from spinner.DIALix.COM (localhost.DIALix.oz.au [127.0.0.1]) by spinner.DIALix.COM (8.8.4/8.8.4) with ESMTP id QAA24233; Sun, 16 Feb 1997 16:21:45 +0800 (WST) Message-Id: <199702160821.QAA24233@spinner.DIALix.COM> X-Mailer: exmh version 2.0gamma 1/27/96 To: Mike Pritchard cc: bde@zeta.org.au (Bruce Evans), ache@freefall.freebsd.org, cvs-all@freefall.freebsd.org, CVS-committers@freefall.freebsd.org, cvs-sys@freefall.freebsd.org Subject: Re: cvs commit: src/sys/sys types.h In-reply-to: Your message of "Sat, 15 Feb 1997 23:02:18 PST." <199702160702.XAA16636@freefall.freebsd.org> Date: Sun, 16 Feb 1997 16:21:44 +0800 From: Peter Wemm Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Mike Pritchard wrote: > Bruce Evans wrote: > > > > >ache 97/02/15 14:26:30 > > > > > > Modified: sys/sys types.h > > > Log: > > > Bump default FD_SETSIZE from 256 to 1024 as many modern > > > systems do nowdays (like SunOs 5.5.1 f.e.) > > > 256 is too small under real network load > > > > Please back it out. The kernel is not read for this. It always rounds up > > to a multiple of FD_SETSIZE bits. This will clobber old applications. > > It also bzeros a multiple of 6 * FDSETSIZE bits. This will take a > > fairly long time with a large FD_SETSIZE. It already takes too long. > > All this is because select() is specified to handle `struct fd_set's. > > The kernel handles whole objects. > > Sounds like it is time for us to implement the poll system > call. I noticed that OpenBSD has already done so. So have I.. But then again, you're all probably sick of hearing that by now :-] I note that NetBSD have gone further and removed the file/vnode/devsw select entries and replaced them with poll entries. select() and poll() are both system calls, but they both share the same device driver/etc hooks. There is a good reason why NetBSD have done this.. The poll device driver interface is a superset of select's hooks. By implementing poll() in terms of select's kernel/driver backends, you loose a lot of information (poll does much more than testing for read/write/error). But implementing the select syscall with the poll kernel/driver back ends looses no information in either case and is virtually for free. The problem is that it's a device driver interface change. I'd like to follow NetBSD here, the major changes in 3.0 (eg: lite2, smp, etc) are probably the best chance we'll get to do it while containing the chaos in one single jump. It's not application visible except that a poll implementation would be much more accurate and useful. Since select/poll backends go through the fileops and VOP tables, it'd be relatively simple to be able to poll() a directory file so that you wake up when a file is created or deleted in it, or to poll() a file for a size change etc. This requires the poll backend - the select backend only does read/write polling and cannot easily be extended. The select backends are defined to take a "which" argument of 0 (exception/error), FREAD or FWRITE. A poll backend takes a "which" argument that is interpreted as a bit mask so that you can do fancy stuff like select for readable urgent data, not just "readable" then try to determine whether it was urgent or not. To implement polling on state/size changes on files or directories, ufs_poll() would replace ufs_select() (currently returns true), and would actually do something useful. Also, bzero(6*FD_SETSIZE...) etc can probably be fixed without too much pain for select() if somebody's prepared to get blood on their hands. The incoming bits do not appear to need to be zeroed (copyin overwrites it with a rounded up to byte boundary and counts the bits that it's scanning in selscan), and only the number of bytes that are going to be copyout'ed need to be zeroed. Also, aren't you misreading the roundup? /* The amount of space we need to allocate */ ni = howmany(roundup2 (uap->nd, FD_SETSIZE), NFDBITS) * sizeof(fd_mask); if (ni > p->p_selbits_size) { /* realloc p_selbits */ } ...later... /* The amount of space we need to copyin/copyout */ ni = howmany(uap->nd, NFDBITS) * sizeof(fd_mask); #define getbits(name, x) \ if (uap->name && \ (error = copyin((caddr_t)uap->name, (caddr_t)ibits[x], ni))) \ goto done; getbits(in, 0); getbits(ou, 1); getbits(ex, 2); Note the reuse of 'ni'. The application is only exposed to sizeof(fd_mask) (fd_mask == unsigned long) rounding, not FD_SETSIZE. The FD_SETSIZE rounding is only to size the initial workspace to save excess reallocation calls to the kernel allocator. Even then it expands by 32 bytes (256 bits), not FD_SETSIZE. Also, it seems to me that the code is not very optimal. I'm probably about to embarress myself here, but wouldn't it be better something like this?: Change: /* * This buffer is usually very small therefore it's probably faster * to just zero it, rather than calculate what needs to be zeroed. */ bzero (p->p_selbits, p->p_selbits_size * 6); /* The amount of space we need to copyin/copyout */ ni = howmany(uap->nd, NFDBITS) * sizeof(fd_mask); to something more like this: /* The amount of space we need to copyin/copyout */ ni = howmany(uap->nd, NFDBITS) * sizeof(fd_mask); if (ni > 8) { /* 64 bits - arbitary guess at tradeoff */ /* * Clear the entire return-to-user buffer space */ bzero (p->p_selbits + (i + 3) * p->p_selbits_size, p->p_selbits_size * 3); } else { for (i = 0; i < 3; i++) bzero(obits[i], ni); } The idea was to figure out where the cost of three function calls versus unnecessary clearing of chunks of memory pays off. Urk... Enough rambling from me.. > > Bruce > > > > > -- > Mike Pritchard > mpp@FreeBSD.org > "Go that way. Really fast. If something gets in your way, turn" Cheers, -Peter From owner-cvs-sys Sun Feb 16 05:22:59 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id FAA04180 for cvs-sys-outgoing; Sun, 16 Feb 1997 05:22:59 -0800 (PST) Received: from sovcom.kiae.su (sovcom.kiae.su [193.125.152.1]) by freefall.freebsd.org (8.8.5/8.8.5) with SMTP id FAA04120; Sun, 16 Feb 1997 05:21:05 -0800 (PST) Received: by sovcom.kiae.su id AA13904 (5.65.kiae-1 ); Sun, 16 Feb 1997 16:05:34 +0300 Received: by sovcom.KIAE.su (UUMAIL/2.0); Sun, 16 Feb 97 16:05:33 +0300 Received: (from ache@localhost) by nagual.ru (8.8.5/8.8.5) id QAA00485; Sun, 16 Feb 1997 16:04:07 +0300 (MSK) Date: Sun, 16 Feb 1997 16:04:02 +0300 (MSK) From: =?KOI8-R?B?4c7E0sXKIP7F0s7P1w==?= To: Peter Wemm Cc: Mike Pritchard , Bruce Evans , cvs-all@freefall.freebsd.org, CVS-committers@freefall.freebsd.org, cvs-sys@freefall.freebsd.org Subject: Re: cvs commit: src/sys/sys types.h In-Reply-To: <199702160821.QAA24233@spinner.DIALix.COM> Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Sun, 16 Feb 1997, Peter Wemm wrote: > poll... Well lets fix select for now and take poll as next step. > Also, bzero(6*FD_SETSIZE...) etc can probably be fixed without too much > pain for select() if somebody's prepared to get blood on their hands. The > incoming bits do not appear to need to be zeroed (copyin overwrites it > with a rounded up to byte boundary and counts the bits that it's scanning > in selscan), and only the number of bytes that are going to be copyout'ed > need to be zeroed. I don't understand why rounding is even needed here (maybe to make less overhead for p_selbits_size increasing?), but we can always round to 256 boundary by defining FD_ROUNDSIZE=256 in sys_generic.c > if (ni > 8) { /* 64 bits - arbitary guess at tradeoff */ > /* > * Clear the entire return-to-user buffer space > */ > bzero (p->p_selbits + (i + 3) * p->p_selbits_size, ^^^^^^ just 3 > p->p_selbits_size * 3); I agree about this one too and include it the patch. Here is the patch variant for review: *** sys_generic.c.orig Tue Jan 14 23:11:28 1997 --- sys_generic.c Sun Feb 16 16:00:48 1997 *************** *** 508,513 **** --- 508,515 ---- return (error); } + #define FD_ROUNDSIZE 256 + static int nselcoll; int selwait; *************** *** 539,545 **** uap->nd = p->p_fd->fd_nfiles; /* forgiving; slightly wrong */ /* The amount of space we need to allocate */ ! ni = howmany(roundup2 (uap->nd, FD_SETSIZE), NFDBITS) * sizeof(fd_mask); if (ni > p->p_selbits_size) { --- 541,547 ---- uap->nd = p->p_fd->fd_nfiles; /* forgiving; slightly wrong */ /* The amount of space we need to allocate */ ! ni = howmany(roundup2 (uap->nd, FD_ROUNDSIZE), NFDBITS) * sizeof(fd_mask); if (ni > p->p_selbits_size) { *************** *** 547,553 **** free (p->p_selbits, M_SELECT); while (p->p_selbits_size < ni) ! p->p_selbits_size += 32; /* Increase by 256 bits */ p->p_selbits = malloc(p->p_selbits_size * 6, M_SELECT, M_WAITOK); --- 549,557 ---- free (p->p_selbits, M_SELECT); while (p->p_selbits_size < ni) ! p->p_selbits_size += ! howmany(FD_ROUNDSIZE, NFDBITS) * ! sizeof(fd_mask); p->p_selbits = malloc(p->p_selbits_size * 6, M_SELECT, M_WAITOK); *************** *** 558,571 **** p->p_selbits_size); } - /* - * This buffer is usually very small therefore it's probably faster - * to just zero it, rather than calculate what needs to be zeroed. - */ - bzero (p->p_selbits, p->p_selbits_size * 6); - /* The amount of space we need to copyin/copyout */ ni = howmany(uap->nd, NFDBITS) * sizeof(fd_mask); #define getbits(name, x) \ if (uap->name && \ --- 562,580 ---- p->p_selbits_size); } /* The amount of space we need to copyin/copyout */ ni = howmany(uap->nd, NFDBITS) * sizeof(fd_mask); + + if (ni > 8) { /* 64 bits - arbitary guess at tradeoff */ + /* + * Clear the entire return-to-user buffer space + */ + bzero (p->p_selbits + 3 * p->p_selbits_size, + p->p_selbits_size * 3); + } else { + for (i = 0; i < 3; i++) + bzero(obits[i], ni); + } #define getbits(name, x) \ if (uap->name && \ -- Andrey A. Chernov http://www.nagual.ru/~ache/ From owner-cvs-sys Sun Feb 16 05:32:28 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id FAA04607 for cvs-sys-outgoing; Sun, 16 Feb 1997 05:32:28 -0800 (PST) Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id FAA04592; Sun, 16 Feb 1997 05:32:15 -0800 (PST) Received: (from bde@localhost) by godzilla.zeta.org.au (8.8.3/8.6.9) id AAA19640; Mon, 17 Feb 1997 00:28:19 +1100 Date: Mon, 17 Feb 1997 00:28:19 +1100 From: Bruce Evans Message-Id: <199702161328.AAA19640@godzilla.zeta.org.au> To: mpp@freefall.freebsd.org, peter@spinner.dialix.com Subject: Re: cvs commit: src/sys/sys types.h Cc: ache@freefall.freebsd.org, bde@zeta.org.au, cvs-all@freefall.freebsd.org, CVS-committers@freefall.freebsd.org, cvs-sys@freefall.freebsd.org Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >> Sounds like it is time for us to implement the poll system >> call. I noticed that OpenBSD has already done so. > >So have I.. But then again, you're all probably sick of hearing >that by now :-] How much slower would it be? :-) >Also, aren't you misreading the roundup? > > /* The amount of space we need to allocate */ > ni = howmany(roundup2 (uap->nd, FD_SETSIZE), NFDBITS) * > sizeof(fd_mask); > if (ni > p->p_selbits_size) { > /* realloc p_selbits */ > } > ...later... > /* The amount of space we need to copyin/copyout */ > ni = howmany(uap->nd, NFDBITS) * sizeof(fd_mask); Oops. I had forgotten that select() doesn't actually initialize the entire descriptor set like its man page says. It only zeros a convenient number of bits above `nfds'. We considered only rounding to a byte boundary but decided to be bug for bug compatible. The first `ni' seems to be left over from a version that wanted to round to an fd_set boundary. It doesn't hurt to allocate a few extra bits, but bzeroing them is just wasteful. >Note the reuse of 'ni'. The application is only exposed to sizeof(fd_mask) >(fd_mask == unsigned long) rounding, not FD_SETSIZE. The FD_SETSIZE >rounding is only to size the initial workspace to save excess reallocation >calls to the kernel allocator. Even then it expands by 32 bytes (256 >bits), not FD_SETSIZE. Yes, the change for user space was OK except for performance problems. Sorry for the false alarm. >Also, it seems to me that the code is not very optimal. I'm probably >about to embarress myself here, but wouldn't it be better something like >this?: >... It should just clear 6 * (minimum (rounded) number of bits) and not use the bits at the end. Bruce From owner-cvs-sys Sun Feb 16 06:06:33 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id GAA05813 for cvs-sys-outgoing; Sun, 16 Feb 1997 06:06:33 -0800 (PST) Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id GAA05807; Sun, 16 Feb 1997 06:06:26 -0800 (PST) Received: (from bde@localhost) by godzilla.zeta.org.au (8.8.3/8.6.9) id BAA20251; Mon, 17 Feb 1997 01:04:24 +1100 Date: Mon, 17 Feb 1997 01:04:24 +1100 From: Bruce Evans Message-Id: <199702161404.BAA20251@godzilla.zeta.org.au> To: ache@nagual.ru, peter@spinner.dialix.com Subject: Re: cvs commit: src/sys/sys types.h Cc: bde@zeta.org.au, cvs-all@freefall.freebsd.org, CVS-committers@freefall.freebsd.org, cvs-sys@freefall.freebsd.org, mpp@freefall.freebsd.org Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >I don't understand why rounding is even needed here (maybe to make less 32-bit rounding is better for bzero, copyout, etc, and doesn't cost more storage. 256-bit rounding avoids slow startup but isn't space-efficient since we're allocating 6 times that. Anyway, I think it would be better to use an auto buffer for small allocations even p_selbits has been allocated. This would have better locality. Most applications wouldn't need a malloced buffer. Half of implementation: fd_mask selbits[6 * OLD_FD_SETSIZE / NFDBITS]; ... ni = bits_to_copyin(); /* multiple of NFDBITS */ if (ni <= OLD_FD_SETSIZE) use_selbits(); else use_malloced_bits(); >*** sys_generic.c.orig Tue Jan 14 23:11:28 1997 >--- sys_generic.c Sun Feb 16 16:00:48 1997 >*************** >*** 508,513 **** >--- 508,515 ---- > return (error); > } > >+ #define FD_ROUNDSIZE 256 >+ Change to something like (2048 / 6) = 341. Now that's an odd size :-). >--- 562,580 ---- > p->p_selbits_size); > } > > /* The amount of space we need to copyin/copyout */ > ni = howmany(uap->nd, NFDBITS) * sizeof(fd_mask); >+ >+ if (ni > 8) { /* 64 bits - arbitary guess at tradeoff */ >+ /* >+ * Clear the entire return-to-user buffer space >+ */ >+ bzero (p->p_selbits + 3 * p->p_selbits_size, >+ p->p_selbits_size * 3); >+ } else { >+ for (i = 0; i < 3; i++) >+ bzero(obits[i], ni); >+ } Don't use this except possibly for very large ni together with unusual combinations of fds. Put 3 contiguous copyout buffers first (readfds, writefds, exexptfds), then 3 copyin buffers, and bzero 1*ni, 2*ni or 3*ni bits as necessary (exceptfds is usually NULL so the bzero doesn't need to cover it). The old code got this wrong mainly by spacing the buffers with separation p->p_selbits_size instead of ni. Bruce From owner-cvs-sys Sun Feb 16 09:26:10 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id JAA14040 for cvs-sys-outgoing; Sun, 16 Feb 1997 09:26:10 -0800 (PST) Received: from jekyll.piermont.com (jekyll.piermont.com [206.1.51.15]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id JAA14029; Sun, 16 Feb 1997 09:25:54 -0800 (PST) Received: from [[UNIX: localhost]] ([[UNIX: localhost]]) by jekyll.piermont.com (8.8.5/8.6.12) with SMTP id MAA16517; Sun, 16 Feb 1997 12:25:52 -0500 (EST) Message-Id: <199702161725.MAA16517@jekyll.piermont.com> X-Authentication-Warning: jekyll.piermont.com: [[UNIX: localhost]] didn't use HELO protocol To: Mike Pritchard cc: bde@zeta.org.au (Bruce Evans), ache@freefall.freebsd.org, cvs-all@freefall.freebsd.org, CVS-committers@freefall.freebsd.org, cvs-sys@freefall.freebsd.org Subject: Re: cvs commit: src/sys/sys types.h In-reply-to: Your message of "Sat, 15 Feb 1997 23:02:18 PST." <199702160702.XAA16636@freefall.freebsd.org> Reply-To: perry@piermont.com X-Reposting-Policy: redistribute only with permission Date: Sun, 16 Feb 1997 12:25:46 -0500 From: "Perry E. Metzger" Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Mike Pritchard writes: > Sounds like it is time for us to implement the poll system > call. I noticed that OpenBSD has already done so. NetBSD implemented it. OpenBSD probably has it as a result. > Bruce Evans wrote: > > > > >ache 97/02/15 14:26:30 > > > > > > Modified: sys/sys types.h > > > Log: > > > Bump default FD_SETSIZE from 256 to 1024 as many modern > > > systems do nowdays (like SunOs 5.5.1 f.e.) > > > 256 is too small under real network load > > > > Please back it out. The kernel is not read for this. It always rounds up > > to a multiple of FD_SETSIZE bits. This will clobber old applications. > > It also bzeros a multiple of 6 * FDSETSIZE bits. This will take a > > fairly long time with a large FD_SETSIZE. It already takes too long. > > All this is because select() is specified to handle `struct fd_set's. > > The kernel handles whole objects. > > > > Bruce > > > > > -- > Mike Pritchard > mpp@FreeBSD.org > "Go that way. Really fast. If something gets in your way, turn" From owner-cvs-sys Sun Feb 16 10:26:38 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id KAA17823 for cvs-sys-outgoing; Sun, 16 Feb 1997 10:26:38 -0800 (PST) Received: (from bde@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id KAA17807; Sun, 16 Feb 1997 10:26:35 -0800 (PST) Date: Sun, 16 Feb 1997 10:26:35 -0800 (PST) From: Bruce Evans Message-Id: <199702161826.KAA17807@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-lib, cvs-sys, cvs-etc Subject: cvs commit: src/lib/msun Makefile src/etc make.conf src/lib/msun/src get_hw_float.c e_acos.c e_asin.c e_atan2.c e_exp.c e_fmod.c e_log.c e_log10.c e_remainder.c e_scalb.c e_sqrt.c s_atan.c s_ceil.c s_copysign.c s_cos.c s_finite.c s_floor.c s_ilogb.c s_logb.c s_rint.c s_scalbn.c s_significand.c s_sin.c s_tan.c src/sys/i386/include asmacros.h Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk bde 97/02/16 10:26:33 Modified: etc make.conf lib/msun Makefile lib/msun/src e_acos.c e_asin.c e_atan2.c e_exp.c e_fmod.c e_log.c e_log10.c e_remainder.c e_scalb.c e_sqrt.c s_atan.c s_ceil.c s_copysign.c s_cos.c s_finite.c s_floor.c s_ilogb.c s_logb.c s_rint.c s_scalbn.c s_significand.c s_sin.c s_tan.c sys/i386/include asmacros.h Added: lib/msun/src get_hw_float.c Log: Select between the generic math functions and the i387-specific ones at runtime. etc/make.conf: Nuked HAVE_FPU option. lib/msun/Makefile: Always build the i387 objects. Copy the i387 source files at build time so that the i387 objects have different names. This is simpler than renaming the files in the cvs repository or repeating half of bsd.lib.mk to add explicit rules. lib/msun/src/*.c: Renamed all functions that have an i387-specific version by adding `__generic_' to their names. lib/msun/src/get_hw_float.c: New file for getting machdep.hw_float from the kernel. sys/i386/include/asmacros.h: Abuse the ENTRY() macro to generate jump vectors and associated code. This works much like PIC PLT dynamic initialization. The PIC case is messy. The old i387 entry points are renamed. Renaming is easier here because the names are given by macro expansions. Revision Changes Path 1.42 +0 -6 src/etc/make.conf 1.16 +13 -9 src/lib/msun/Makefile 1.4 +2 -2 src/lib/msun/src/e_acos.c 1.5 +2 -2 src/lib/msun/src/e_asin.c 1.5 +2 -2 src/lib/msun/src/e_atan2.c 1.5 +2 -2 src/lib/msun/src/e_exp.c 1.4 +2 -2 src/lib/msun/src/e_fmod.c 1.4 +2 -2 src/lib/msun/src/e_log.c 1.5 +2 -2 src/lib/msun/src/e_log10.c 1.4 +2 -2 src/lib/msun/src/e_remainder.c 1.4 +4 -4 src/lib/msun/src/e_scalb.c 1.4 +2 -2 src/lib/msun/src/e_sqrt.c 1.4 +2 -2 src/lib/msun/src/s_atan.c 1.4 +2 -2 src/lib/msun/src/s_ceil.c 1.4 +2 -2 src/lib/msun/src/s_copysign.c 1.4 +2 -2 src/lib/msun/src/s_cos.c 1.4 +2 -2 src/lib/msun/src/s_finite.c 1.4 +2 -2 src/lib/msun/src/s_floor.c 1.4 +2 -2 src/lib/msun/src/s_ilogb.c 1.4 +2 -2 src/lib/msun/src/s_logb.c 1.5 +2 -2 src/lib/msun/src/s_rint.c 1.4 +2 -2 src/lib/msun/src/s_scalbn.c 1.4 +2 -2 src/lib/msun/src/s_significand.c 1.4 +2 -2 src/lib/msun/src/s_sin.c 1.4 +2 -2 src/lib/msun/src/s_tan.c 1.12 +77 -0 src/sys/i386/include/asmacros.h From owner-cvs-sys Sun Feb 16 14:32:49 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id OAA02103 for cvs-sys-outgoing; Sun, 16 Feb 1997 14:32:49 -0800 (PST) Received: from time.cdrom.com (time.cdrom.com [204.216.27.226]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id OAA02046; Sun, 16 Feb 1997 14:32:26 -0800 (PST) Received: from time.cdrom.com (localhost [127.0.0.1]) by time.cdrom.com (8.8.5/8.6.9) with ESMTP id OAA15474; Sun, 16 Feb 1997 14:32:24 -0800 (PST) To: Bruce Evans cc: CVS-committers@freefall.freebsd.org, cvs-all@freefall.freebsd.org, cvs-lib@freefall.freebsd.org, cvs-sys@freefall.freebsd.org, cvs-etc@freefall.freebsd.org Subject: Re: cvs commit: src/lib/msun Makefile src/etc make.conf src/lib/msun/src get_hw_float.c e_acos.c e_asin.c e_atan2.c e_exp.c e_fmod.c e_log.c e_log10.c e_remainder.c e_scalb.c e_sqrt.c s_atan.c s_ceil.c s_copysign.c s_cos.c s_finite.c s_floor.c s_ilogb.c s_logb.c s_rint.c s_scalbn.c s_significand.c s_sin.c s_tan.c src/sys/i386/include asmacros.h In-reply-to: Your message of "Sun, 16 Feb 1997 10:26:35 PST." <199702161826.KAA17807@freefall.freebsd.org> Date: Sun, 16 Feb 1997 14:32:24 -0800 Message-ID: <15471.856132344@time.cdrom.com> From: "Jordan K. Hubbard" Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > etc/make.conf: > Nuked HAVE_FPU option. And there was wild applause! :-) Jordan From owner-cvs-sys Sun Feb 16 14:58:04 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id OAA03455 for cvs-sys-outgoing; Sun, 16 Feb 1997 14:58:04 -0800 (PST) Received: from sovcom.kiae.su (sovcom.kiae.su [193.125.152.1]) by freefall.freebsd.org (8.8.5/8.8.5) with SMTP id OAA03444; Sun, 16 Feb 1997 14:57:55 -0800 (PST) Received: by sovcom.kiae.su id AA18982 (5.65.kiae-1 ); Mon, 17 Feb 1997 01:38:35 +0300 Received: by sovcom.KIAE.su (UUMAIL/2.0); Mon, 17 Feb 97 01:38:35 +0300 Received: (from ache@localhost) by nagual.ru (8.8.5/8.8.5) id BAA00738; Mon, 17 Feb 1997 01:37:40 +0300 (MSK) Date: Mon, 17 Feb 1997 01:37:35 +0300 (MSK) From: =?KOI8-R?B?4c7E0sXKIP7F0s7P1w==?= To: Bruce Evans Cc: peter@spinner.dialix.com, cvs-all@freefall.freebsd.org, CVS-committers@freefall.freebsd.org, cvs-sys@freefall.freebsd.org, mpp@freefall.freebsd.org Subject: Re: cvs commit: src/sys/sys types.h In-Reply-To: <199702161404.BAA20251@godzilla.zeta.org.au> Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Mon, 17 Feb 1997, Bruce Evans wrote: > >+ #define FD_ROUNDSIZE 256 > >+ > > Change to something like (2048 / 6) = 341. Now that's an odd size :-). I change it (see update patch below) but still not understand why 2048 is here. p_selbits_size increasing constant will be 44 in this case, is it good value for you too? > Don't use this except possibly for very large ni together with unusual > combinations of fds. Put 3 contiguous copyout buffers first (readfds, > writefds, exexptfds), then 3 copyin buffers, and bzero 1*ni, 2*ni or > 3*ni bits as necessary (exceptfds is usually NULL so the bzero doesn't > need to cover it). The old code got this wrong mainly by spacing the > buffers with separation p->p_selbits_size instead of ni. I try to implement what you say, see the patch below: *** sys_generic.c.orig Tue Jan 14 23:11:28 1997 --- sys_generic.c Mon Feb 17 01:33:11 1997 *************** *** 508,513 **** --- 508,515 ---- return (error); } + #define FD_ROUNDSIZE (2048 / (3 + 3)) + static int nselcoll; int selwait; *************** *** 539,545 **** uap->nd = p->p_fd->fd_nfiles; /* forgiving; slightly wrong */ /* The amount of space we need to allocate */ ! ni = howmany(roundup2 (uap->nd, FD_SETSIZE), NFDBITS) * sizeof(fd_mask); if (ni > p->p_selbits_size) { --- 541,547 ---- uap->nd = p->p_fd->fd_nfiles; /* forgiving; slightly wrong */ /* The amount of space we need to allocate */ ! ni = howmany(roundup2 (uap->nd, FD_ROUNDSIZE), NFDBITS) * sizeof(fd_mask); if (ni > p->p_selbits_size) { *************** *** 547,571 **** free (p->p_selbits, M_SELECT); while (p->p_selbits_size < ni) ! p->p_selbits_size += 32; /* Increase by 256 bits */ p->p_selbits = malloc(p->p_selbits_size * 6, M_SELECT, M_WAITOK); } - for (i = 0; i < 3; i++) { - ibits[i] = (fd_mask *)(p->p_selbits + i * p->p_selbits_size); - obits[i] = (fd_mask *)(p->p_selbits + (i + 3) * - p->p_selbits_size); - } - - /* - * This buffer is usually very small therefore it's probably faster - * to just zero it, rather than calculate what needs to be zeroed. - */ - bzero (p->p_selbits, p->p_selbits_size * 6); /* The amount of space we need to copyin/copyout */ ni = howmany(uap->nd, NFDBITS) * sizeof(fd_mask); #define getbits(name, x) \ if (uap->name && \ --- 549,579 ---- free (p->p_selbits, M_SELECT); while (p->p_selbits_size < ni) ! p->p_selbits_size += ! howmany(FD_ROUNDSIZE, NFDBITS) * ! sizeof(fd_mask); p->p_selbits = malloc(p->p_selbits_size * 6, M_SELECT, M_WAITOK); } /* The amount of space we need to copyin/copyout */ ni = howmany(uap->nd, NFDBITS) * sizeof(fd_mask); + + for (i = 0; i < 3; i++) { + obits[i] = (fd_mask *)(p->p_selbits + i * ni); + ibits[i] = (fd_mask *)(p->p_selbits + (i + 3) * ni); + } + if (uap->ex) + i = 3; + else if (uap->ou) + i = 2; + else if (uap->in) + i = 1; + else + i = 0; + if (i) + bzero(p->p_selbits, i * ni); #define getbits(name, x) \ if (uap->name && \ -- Andrey A. Chernov http://www.nagual.ru/~ache/ From owner-cvs-sys Sun Feb 16 16:26:26 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id QAA09142 for cvs-sys-outgoing; Sun, 16 Feb 1997 16:26:26 -0800 (PST) Received: from lestat.nas.nasa.gov (lestat.nas.nasa.gov [129.99.50.29]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id QAA09130; Sun, 16 Feb 1997 16:26:18 -0800 (PST) Received: from localhost (localhost [127.0.0.1]) by lestat.nas.nasa.gov (8.7.5/8.6.12) with SMTP id QAA27668; Sun, 16 Feb 1997 16:22:42 -0800 (PST) Message-Id: <199702170022.QAA27668@lestat.nas.nasa.gov> X-Authentication-Warning: lestat.nas.nasa.gov: Host localhost [127.0.0.1] didn't use HELO protocol To: Bruce Evans Cc: mpp@freefall.freebsd.org, peter@spinner.dialix.com, ache@freefall.freebsd.org, cvs-all@freefall.freebsd.org, CVS-committers@freefall.freebsd.org, cvs-sys@freefall.freebsd.org Subject: Re: cvs commit: src/sys/sys types.h Reply-To: Jason Thorpe From: Jason Thorpe Date: Sun, 16 Feb 1997 16:22:42 -0800 Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Mon, 17 Feb 1997 00:28:19 +1100 Bruce Evans wrote: > >> Sounds like it is time for us to implement the poll system > >> call. I noticed that OpenBSD has already done so. > > > >So have I.. But then again, you're all probably sick of hearing > >that by now :-] > > How much slower would it be? :-) You might want to take a look at NetBSD's poll(2) implementation. Note that to really do poll correctly (i.e. not have the same limitations as select(2)), you need to change the interface to the device drivers that implement select entry points. Jason R. Thorpe thorpej@nas.nasa.gov NASA Ames Research Center Home: 408.866.1912 NAS: M/S 258-6 Work: 415.604.0935 Moffett Field, CA 94035 Pager: 415.428.6939 From owner-cvs-sys Sun Feb 16 16:50:35 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id QAA09952 for cvs-sys-outgoing; Sun, 16 Feb 1997 16:50:35 -0800 (PST) Received: from lestat.nas.nasa.gov (lestat.nas.nasa.gov [129.99.50.29]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id QAA09933; Sun, 16 Feb 1997 16:50:20 -0800 (PST) Received: from localhost (localhost [127.0.0.1]) by lestat.nas.nasa.gov (8.7.5/8.6.12) with SMTP id QAA28052; Sun, 16 Feb 1997 16:46:43 -0800 (PST) Message-Id: <199702170046.QAA28052@lestat.nas.nasa.gov> X-Authentication-Warning: lestat.nas.nasa.gov: Host localhost [127.0.0.1] didn't use HELO protocol To: perry@piermont.com Cc: Mike Pritchard , bde@zeta.org.au (Bruce Evans), ache@freefall.freebsd.org, cvs-all@freefall.freebsd.org, CVS-committers@freefall.freebsd.org, cvs-sys@freefall.freebsd.org Subject: Re: cvs commit: src/sys/sys types.h Reply-To: Jason Thorpe From: Jason Thorpe Date: Sun, 16 Feb 1997 16:46:42 -0800 Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Sun, 16 Feb 1997 12:25:46 -0500 "Perry E. Metzger" wrote: > Mike Pritchard writes: > > Sounds like it is time for us to implement the poll system > > call. I noticed that OpenBSD has already done so. > > NetBSD implemented it. OpenBSD probably has it as a result. Actually, this is not the case... OpenBSD implemeted it by using the poll system call emulation from the svr4 compat code (at least, when I looked at the sources when they first got poll, that's how it appeared). NetBSD's poll(2) implementation is completely different from OpenBSD's, and, IMO, better since it's possible to actually implement all of the poll events in the NetBSD implemetnation... To do this, you have to change the interface to the device drivers, which is what NetBSD did. Jason R. Thorpe thorpej@nas.nasa.gov NASA Ames Research Center Home: 408.866.1912 NAS: M/S 258-6 Work: 415.604.0935 Moffett Field, CA 94035 Pager: 415.428.6939 From owner-cvs-sys Sun Feb 16 20:34:22 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id UAA21621 for cvs-sys-outgoing; Sun, 16 Feb 1997 20:34:22 -0800 (PST) Received: from nyx.pr.mcs.net (nyx.pr.mcs.net [204.95.55.81]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id UAA21608; Sun, 16 Feb 1997 20:34:05 -0800 (PST) Received: from nyx.pr.mcs.net (localhost [127.0.0.1]) by nyx.pr.mcs.net (8.8.5/8.8.4) with ESMTP id UAA02370; Sun, 16 Feb 1997 20:08:03 -0600 (CST) Message-Id: <199702170208.UAA02370@nyx.pr.mcs.net> X-Mailer: exmh version 2.0beta 12/23/96 To: Jason Thorpe cc: perry@piermont.com, Mike Pritchard , bde@zeta.org.au (Bruce Evans), ache@freefall.freebsd.org, cvs-all@freefall.freebsd.org, CVS-committers@freefall.freebsd.org, cvs-sys@freefall.freebsd.org Subject: Re: cvs commit: src/sys/sys types.h In-reply-to: Your message of Sun, 16 Feb 1997 16:46:42 -0800. <199702170046.QAA28052@lestat.nas.nasa.gov> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sun, 16 Feb 1997 20:08:03 -0600 From: Chris Csanady Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >NetBSD's poll(2) implementation is completely different from OpenBSD's, >and, IMO, better since it's possible to actually implement all of the >poll events in the NetBSD implemetnation... > >To do this, you have to change the interface to the device drivers, >which is what NetBSD did. If this is to be implemented in freebsd, and changes to the device driver interface are necessary, perhaps now is a good time to try to work toward a standard interface for all the BSD's? I think that this would be something that would be really nice to have, and would be well worth the work to all involoved. I suppose the easiest way for this to happen would be if the NetBSD interface, was acceptable to the FreeBSD team. It might also be a good step in acheiving architecture independance. Any thoughts? Chris Csanady > >Jason R. Thorpe thorpej@nas.nasa.gov >NASA Ames Research Center Home: 408.866.1912 >NAS: M/S 258-6 Work: 415.604.0935 >Moffett Field, CA 94035 Pager: 415.428.6939 From owner-cvs-sys Sun Feb 16 22:44:44 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id WAA27777 for cvs-sys-outgoing; Sun, 16 Feb 1997 22:44:44 -0800 (PST) Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id WAA27770; Sun, 16 Feb 1997 22:44:38 -0800 (PST) Received: (from bde@localhost) by godzilla.zeta.org.au (8.8.3/8.6.9) id RAA18161; Mon, 17 Feb 1997 17:39:50 +1100 Date: Mon, 17 Feb 1997 17:39:50 +1100 From: Bruce Evans Message-Id: <199702170639.RAA18161@godzilla.zeta.org.au> To: ache@nagual.ru, bde@zeta.org.au Subject: Re: cvs commit: src/sys/sys types.h Cc: cvs-all@freefall.freebsd.org, CVS-committers@freefall.freebsd.org, cvs-sys@freefall.freebsd.org, mpp@freefall.freebsd.org, peter@spinner.dialix.com Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >> >+ #define FD_ROUNDSIZE 256 >> >+ >> >> Change to something like (2048 / 6) = 341. Now that's an odd size :-). > >I change it (see update patch below) but still not understand >why 2048 is here. p_selbits_size increasing constant will be 44 in this >case, is it good value for you too? It is really the allocation unit size. 2048 is almost arbitrary. It is the next power of 2 up from the number of bits required for 6 fd_set's with the current FD_SETSIZE (6 * 256 = 1536). p_selbits_size would actually be 256/6 = 42. Although this isn't a multiple of sizeof(fdmask), there is no problem since buffers are spaced every ni bits, not every p->p_selbits_size bits. p_selbits_size should be changed to the size allocated anyway. >*** sys_generic.c.orig Tue Jan 14 23:11:28 1997 >--- sys_generic.c Mon Feb 17 01:33:11 1997 ... >*************** >*** 539,545 **** > uap->nd = p->p_fd->fd_nfiles; /* forgiving; slightly wrong */ > > /* The amount of space we need to allocate */ >! ni = howmany(roundup2 (uap->nd, FD_SETSIZE), NFDBITS) * > sizeof(fd_mask); > > if (ni > p->p_selbits_size) { >--- 541,547 ---- > uap->nd = p->p_fd->fd_nfiles; /* forgiving; slightly wrong */ > > /* The amount of space we need to allocate */ >! ni = howmany(roundup2 (uap->nd, FD_ROUNDSIZE), NFDBITS) * > sizeof(fd_mask); > > if (ni > p->p_selbits_size) { Don't round up to the allocation unit here. Set ni = roundup(uap->nd, NFDBITS). Then if (ni <= (OLD_FD_SETSIZE = 32 * NBBY)), use an auto buffer of size 6*OLD_FD_SETSIZE bits. Otherwise, use a malloced buffer (malloc a larger one if ni > p->p_selbits_size). This can be simplified by always mallocing and freeing at the end. Space can be saved by reducing ni first for the common case when there are no exceptfd's etc. >--- 549,579 ---- > free (p->p_selbits, M_SELECT); > > while (p->p_selbits_size < ni) >! p->p_selbits_size += >! howmany(FD_ROUNDSIZE, NFDBITS) * >! sizeof(fd_mask); This can be simplified to something like p_selbytes_allocated = roundup2(6 * ni, FD_ROUNDSIZE) / NBBY. Don't use exactly this - it assumes that FD_ROUNDSIZE is both a multiple of NBBY and a power of 2. FD_ROUNDSIZE should be defined as a multiple of NBBY, and it won't be a power of 2 on weird machines. > /* The amount of space we need to copyin/copyout */ > ni = howmany(uap->nd, NFDBITS) * sizeof(fd_mask); Simplify this to something ni /= NBBY (maybe rename the variable). >+ >+ for (i = 0; i < 3; i++) { >+ obits[i] = (fd_mask *)(p->p_selbits + i * ni); >+ ibits[i] = (fd_mask *)(p->p_selbits + (i + 3) * ni); >+ } >+ if (uap->ex) >+ i = 3; >+ else if (uap->ou) >+ i = 2; >+ else if (uap->in) >+ i = 1; >+ else >+ i = 0; >+ if (i) >+ bzero(p->p_selbits, i * ni); I think there is a problem caused by stupid handling of NULL fdsets. Nothing is copied in, but selscan() wastes time scanning the buffers. It depends on the current behaviour of bzeroing everything. Fixing this would be a useful optimization than reducing the amount to bzero :-). Bruce From owner-cvs-sys Sun Feb 16 23:21:31 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id XAA28820 for cvs-sys-outgoing; Sun, 16 Feb 1997 23:21:31 -0800 (PST) Received: (from bde@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id XAA28813; Sun, 16 Feb 1997 23:21:30 -0800 (PST) Date: Sun, 16 Feb 1997 23:21:30 -0800 (PST) From: Bruce Evans Message-Id: <199702170721.XAA28813@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/i386/include asmacros.h Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk bde 97/02/16 23:21:29 Modified: sys/i386/include asmacros.h Log: Replaced START_ENTRY by _START_ENTRY. -current hasn't got my cleanup of DEFS.h which renamed it. Revision Changes Path 1.13 +6 -6 src/sys/i386/include/asmacros.h From owner-cvs-sys Mon Feb 17 02:58:08 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id CAA08197 for cvs-sys-outgoing; Mon, 17 Feb 1997 02:58:08 -0800 (PST) Received: (from davidg@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id CAA08190; Mon, 17 Feb 1997 02:58:06 -0800 (PST) Date: Mon, 17 Feb 1997 02:58:06 -0800 (PST) From: David Greenman Message-Id: <199702171058.CAA08190@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/kern kern_fork.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk davidg 97/02/17 02:58:05 Branch: sys/kern RELENG_2_1_0 Modified: sys/kern kern_fork.c Log: Pass P_SUGID on to the child of a fork(). It was possible to get rlogin to coredump previously since it (somewhat uniquely) is setuid and forks without execing, and thus without passing P_SUGID the child could coredump and possibly divulge sensitive information (such as encrypted passwords from the passwd database). Revision Changes Path 1.12.4.3 +5 -1 src/sys/kern/kern_fork.c From owner-cvs-sys Mon Feb 17 02:58:52 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id CAA08283 for cvs-sys-outgoing; Mon, 17 Feb 1997 02:58:52 -0800 (PST) Received: (from davidg@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id CAA08276; Mon, 17 Feb 1997 02:58:49 -0800 (PST) Date: Mon, 17 Feb 1997 02:58:49 -0800 (PST) From: David Greenman Message-Id: <199702171058.CAA08276@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/kern kern_fork.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk davidg 97/02/17 02:58:48 Modified: sys/kern kern_fork.c Log: Pass P_SUGID on to the child of a fork(). It was possible to get rlogin to coredump previously since it (somewhat uniquely) is setuid and forks without execing, and thus without passing P_SUGID the child could coredump and possibly divulge sensitive information (such as encrypted passwords from the passwd database). Revision Changes Path 1.31 +4 -0 src/sys/kern/kern_fork.c From owner-cvs-sys Mon Feb 17 02:59:45 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id CAA08350 for cvs-sys-outgoing; Mon, 17 Feb 1997 02:59:45 -0800 (PST) Received: (from davidg@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id CAA08343; Mon, 17 Feb 1997 02:59:44 -0800 (PST) Date: Mon, 17 Feb 1997 02:59:44 -0800 (PST) From: David Greenman Message-Id: <199702171059.CAA08343@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/kern kern_fork.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk davidg 97/02/17 02:59:43 Branch: sys/kern RELENG_2_2 Modified: sys/kern kern_fork.c Log: Brought in change from rev 1.31: pass P_SUGID on to the child of fork(). Revision Changes Path 1.27.2.2 +5 -1 src/sys/kern/kern_fork.c From owner-cvs-sys Mon Feb 17 12:51:26 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id MAA13732 for cvs-sys-outgoing; Mon, 17 Feb 1997 12:51:26 -0800 (PST) Received: from sovcom.kiae.su (sovcom.kiae.su [193.125.152.1]) by freefall.freebsd.org (8.8.5/8.8.5) with SMTP id MAA13710; Mon, 17 Feb 1997 12:51:17 -0800 (PST) Received: by sovcom.kiae.su id AA00281 (5.65.kiae-1 ); Mon, 17 Feb 1997 23:19:31 +0300 Received: by sovcom.KIAE.su (UUMAIL/2.0); Mon, 17 Feb 97 23:19:31 +0300 Received: (from ache@localhost) by nagual.ru (8.8.5/8.8.5) id XAA01697; Mon, 17 Feb 1997 23:14:36 +0300 (MSK) Date: Mon, 17 Feb 1997 23:14:32 +0300 (MSK) From: =?KOI8-R?B?4c7E0sXKIP7F0s7P1w==?= To: Bruce Evans Cc: cvs-all@freefall.freebsd.org, CVS-committers@freefall.freebsd.org, cvs-sys@freefall.freebsd.org, mpp@freefall.freebsd.org, peter@spinner.dialix.com Subject: Re: cvs commit: src/sys/sys types.h In-Reply-To: <199702170639.RAA18161@godzilla.zeta.org.au> Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Mon, 17 Feb 1997, Bruce Evans wrote: > It is really the allocation unit size. 2048 is almost arbitrary. It is > the next power of 2 up from the number of bits required for 6 fd_set's > with the current FD_SETSIZE (6 * 256 = 1536). Hmm. I don't think that this constant plays any essential role, the only requirement that re-allocation must not occurse too often. In my new variant I not use it at all and code looks much cleaner. > Don't round up to the allocation unit here. Set ni = roundup(uap->nd, > NFDBITS). Then if (ni <= (OLD_FD_SETSIZE = 32 * NBBY)), use an auto > buffer of size 6*OLD_FD_SETSIZE bits. Otherwise, use a malloced buffer > (malloc a larger one if ni > p->p_selbits_size). This can be simplified > by always mallocing and freeing at the end. Space can be saved by > reducing ni first for the common case when there are no exceptfd's etc. I implement somehing like you say. > >--- 549,579 ---- > > free (p->p_selbits, M_SELECT); > > > > while (p->p_selbits_size < ni) > >! p->p_selbits_size += > >! howmany(FD_ROUNDSIZE, NFDBITS) * > >! sizeof(fd_mask); > > This can be simplified to something like p_selbytes_allocated = > roundup2(6 * ni, FD_ROUNDSIZE) / NBBY. Don't use exactly this - > it assumes that FD_ROUNDSIZE is both a multiple of NBBY and a > power of 2. FD_ROUNDSIZE should be defined as a multiple of NBBY, > and it won't be a power of 2 on weird machines. As I already say I use much simpler thing here, just roundup(). > I think there is a problem caused by stupid handling of NULL fdsets. > Nothing is copied in, but selscan() wastes time scanning the buffers. > It depends on the current behaviour of bzeroing everything. Fixing > this would be a useful optimization than reducing the amount to > bzero :-). I fix this thing in selscan(). Updated patch variant: *** sys_generic.c.orig Tue Jan 14 23:11:28 1997 --- sys_generic.c Mon Feb 17 22:57:03 1997 *************** *** 508,513 **** --- 508,515 ---- return (error); } + #define FD_CHUNKSIZE 256 + static int nselcoll; int selwait; *************** *** 527,533 **** register struct select_args *uap; int *retval; { ! fd_mask *ibits[3], *obits[3]; struct timeval atv; int s, ncoll, error = 0, timo, i; u_int ni; --- 529,536 ---- register struct select_args *uap; int *retval; { ! fd_mask s_selbits[(3 + 3) * howmany(FD_CHUNKSIZE, NFDBITS)]; ! fd_mask *ibits[3], *obits[3], *selbits; struct timeval atv; int s, ncoll, error = 0, timo, i; u_int ni; *************** *** 538,571 **** if (uap->nd > p->p_fd->fd_nfiles) uap->nd = p->p_fd->fd_nfiles; /* forgiving; slightly wrong */ ! /* The amount of space we need to allocate */ ! ni = howmany(roundup2 (uap->nd, FD_SETSIZE), NFDBITS) * ! sizeof(fd_mask); ! ! if (ni > p->p_selbits_size) { ! if (p->p_selbits_size) ! free (p->p_selbits, M_SELECT); ! while (p->p_selbits_size < ni) ! p->p_selbits_size += 32; /* Increase by 256 bits */ ! p->p_selbits = malloc(p->p_selbits_size * 6, M_SELECT, ! M_WAITOK); } for (i = 0; i < 3; i++) { ! ibits[i] = (fd_mask *)(p->p_selbits + i * p->p_selbits_size); ! obits[i] = (fd_mask *)(p->p_selbits + (i + 3) * ! p->p_selbits_size); } ! ! /* ! * This buffer is usually very small therefore it's probably faster ! * to just zero it, rather than calculate what needs to be zeroed. ! */ ! bzero (p->p_selbits, p->p_selbits_size * 6); ! ! /* The amount of space we need to copyin/copyout */ ! ni = howmany(uap->nd, NFDBITS) * sizeof(fd_mask); #define getbits(name, x) \ if (uap->name && \ --- 541,587 ---- if (uap->nd > p->p_fd->fd_nfiles) uap->nd = p->p_fd->fd_nfiles; /* forgiving; slightly wrong */ ! /* The amount of space we need to copyin/copyout */ ! ni = howmany(uap->nd, NFDBITS) * sizeof(fd_mask); ! if (ni <= sizeof(s_selbits) / (3 + 3)) ! selbits = s_selbits; ! else { ! if (ni > p->p_selbits_size) { ! if (p->p_selbits_size) ! free (p->p_selbits, M_SELECT); ! ! p->p_selbits_size = roundup(ni, ! howmany(FD_CHUNKSIZE, NFDBITS) * ! sizeof(fd_mask)); ! p->p_selbits = malloc(p->p_selbits_size * 6, M_SELECT, ! M_WAITOK); ! } ! selbits = p->p_selbits; } + for (i = 0; i < 3; i++) { ! obits[i] = (fd_mask *)(selbits + i * ni); ! ibits[i] = (fd_mask *)(selbits + (i + 3) * ni); } ! i = 0; ! if (uap->ex) ! i = 3; ! else ! ibits[2] = NULL; ! if (uap->ou) { ! if (i == 0) ! i = 2; ! } else ! ibits[1] = NULL; ! if (uap->in) { ! if (i == 0) ! i = 1; ! } else ! ibits[0] = NULL; ! if (i != 0) ! bzero(selbits, i * ni); #define getbits(name, x) \ if (uap->name && \ *************** *** 654,670 **** static int flag[3] = { FREAD, FWRITE, 0 }; for (msk = 0; msk < 3; msk++) { ! for (i = 0; i < nfd; i += NFDBITS) { ! bits = ibits[msk][i/NFDBITS]; ! while ((j = ffs(bits)) && (fd = i + --j) < nfd) { ! bits &= ~(1 << j); ! fp = fdp->fd_ofiles[fd]; ! if (fp == NULL) ! return (EBADF); ! if ((*fp->f_ops->fo_select)(fp, flag[msk], p)) { ! obits[msk][(fd)/NFDBITS] |= ! (1 << ((fd) % NFDBITS)); ! n++; } } } --- 670,688 ---- static int flag[3] = { FREAD, FWRITE, 0 }; for (msk = 0; msk < 3; msk++) { ! if (ibits[msk]) { ! for (i = 0; i < nfd; i += NFDBITS) { ! bits = ibits[msk][i/NFDBITS]; ! while ((j = ffs(bits)) && (fd = i + --j) < nfd) { ! bits &= ~(1 << j); ! fp = fdp->fd_ofiles[fd]; ! if (fp == NULL) ! return (EBADF); ! if ((*fp->f_ops->fo_select)(fp, flag[msk], p)) { ! obits[msk][(fd)/NFDBITS] |= ! (1 << ((fd) % NFDBITS)); ! n++; ! } } } } -- Andrey A. Chernov http://www.nagual.ru/~ache/ From owner-cvs-sys Mon Feb 17 20:25:40 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id UAA14674 for cvs-sys-outgoing; Mon, 17 Feb 1997 20:25:40 -0800 (PST) Received: (from gibbs@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id UAA14664; Mon, 17 Feb 1997 20:25:37 -0800 (PST) Date: Mon, 17 Feb 1997 20:25:37 -0800 (PST) From: "Justin T. Gibbs" Message-Id: <199702180425.UAA14664@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/i386/scsi aic7xxx.c aic7xxx.h Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk gibbs 97/02/17 20:25:36 Modified: sys/i386/scsi aic7xxx.c aic7xxx.h Log: Don't rely on AAP(Auto Access Pause) when queueing SCBs to the card. This will increase the overhead of queueing a command, but some recent bug reports make me believe that AAP isn't really working and that we are losing some SCBs from the input queue. Hopefully this will cure that problem. Fix some bugs in the error recovery code. Mainly these could cause us to inadvertantly forget to untimeout an SCB that was recovered causing later confusion. Revision Changes Path 1.99 +19 -19 src/sys/i386/scsi/aic7xxx.c 1.38 +1 -1 src/sys/i386/scsi/aic7xxx.h From owner-cvs-sys Mon Feb 17 20:26:57 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id UAA14774 for cvs-sys-outgoing; Mon, 17 Feb 1997 20:26:57 -0800 (PST) Received: (from gibbs@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id UAA14764; Mon, 17 Feb 1997 20:26:55 -0800 (PST) Date: Mon, 17 Feb 1997 20:26:55 -0800 (PST) From: "Justin T. Gibbs" Message-Id: <199702180426.UAA14764@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/i386/scsi aic7xxx.c aic7xxx.h Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk gibbs 97/02/17 20:26:54 Branch: sys/i386/scsi RELENG_2_2 Modified: sys/i386/scsi aic7xxx.c aic7xxx.h Log: Sync with current - Disable use of AAP in queueing commands and fix a few error recovery bugs. Revision Changes Path 1.81.2.9 +19 -19 src/sys/i386/scsi/aic7xxx.c 1.31.2.6 +1 -1 src/sys/i386/scsi/aic7xxx.h From owner-cvs-sys Mon Feb 17 20:27:13 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id UAA14835 for cvs-sys-outgoing; Mon, 17 Feb 1997 20:27:13 -0800 (PST) Received: (from gibbs@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id UAA14828; Mon, 17 Feb 1997 20:27:11 -0800 (PST) Date: Mon, 17 Feb 1997 20:27:11 -0800 (PST) From: "Justin T. Gibbs" Message-Id: <199702180427.UAA14828@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/i386/scsi aic7xxx.c aic7xxx.h Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk gibbs 97/02/17 20:27:10 Branch: sys/i386/scsi RELENG_2_1_0 Modified: sys/i386/scsi aic7xxx.c aic7xxx.h Log: Sync with current - Disable use of AAP in queueing commands and fix a few error recovery bugs. Revision Changes Path 1.29.2.29 +20 -20 src/sys/i386/scsi/aic7xxx.c 1.10.2.13 +2 -2 src/sys/i386/scsi/aic7xxx.h From owner-cvs-sys Mon Feb 17 20:40:45 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id UAA15825 for cvs-sys-outgoing; Mon, 17 Feb 1997 20:40:45 -0800 (PST) Received: (from bde@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id UAA15815; Mon, 17 Feb 1997 20:40:41 -0800 (PST) Date: Mon, 17 Feb 1997 20:40:41 -0800 (PST) From: Bruce Evans Message-Id: <199702180440.UAA15815@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/nfs nfs_vfsops.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk bde 97/02/17 20:40:40 Modified: sys/nfs nfs_vfsops.c Log: Changed `#ifdef COMPAT_PRELITE2' to `#ifndef NO_COMPAT_PRELITE2' so that old nfs mount calls are supported by default. Revision Changes Path 1.34 +5 -5 src/sys/nfs/nfs_vfsops.c From owner-cvs-sys Mon Feb 17 20:58:18 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id UAA16811 for cvs-sys-outgoing; Mon, 17 Feb 1997 20:58:18 -0800 (PST) Received: (from bde@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id UAA16804; Mon, 17 Feb 1997 20:58:16 -0800 (PST) Date: Mon, 17 Feb 1997 20:58:16 -0800 (PST) From: Bruce Evans Message-Id: <199702180458.UAA16804@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/kern vfs_init.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk bde 97/02/17 20:58:16 Modified: sys/kern vfs_init.c Log: Changed `#ifdef COMPAT_PRELITE2' to `#ifndef NO_COMPAT_PRELITE2' so that the old VFS_VFSCONF sysctl is enabled by default. Initialize the vfc_vfsops field to non-NULL in sysctl_ovfs_conf() so that the old VFS_VFSCONF sysctl actually works. The old (still current) getvfsent.c uses this "kernel-only" field to decide which vfs's are configured (the old implementation returned null entries for unconfigured vfs's). Revision Changes Path 1.23 +3 -3 src/sys/kern/vfs_init.c From owner-cvs-sys Mon Feb 17 22:46:45 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id WAA00459 for cvs-sys-outgoing; Mon, 17 Feb 1997 22:46:45 -0800 (PST) Received: (from bde@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id WAA00450; Mon, 17 Feb 1997 22:46:43 -0800 (PST) Date: Mon, 17 Feb 1997 22:46:43 -0800 (PST) From: Bruce Evans Message-Id: <199702180646.WAA00450@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/kern vfs_lookup.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk bde 97/02/17 22:46:42 Modified: sys/kern vfs_lookup.c Log: Fixed namei caching for LOOKUPs. It was broken for lstat() and olstat(). Successful lstat()s purged an existing entry as well as not caching the result. This bug was introduced in Lite1 by setting the LOCKPARENT flag for [o]lstat() in order to support the inherit-attributes-from-parent- directory misfeature for symlinks. LOCKPARENT was previously only set for CREATEs and DELETEs. It is now set for LOOKUPs, but only for [o]lstat(), so the problem wasn't very noticeable. Revision Changes Path 1.15 +2 -1 src/sys/kern/vfs_lookup.c From owner-cvs-sys Mon Feb 17 23:09:31 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id XAA02767 for cvs-sys-outgoing; Mon, 17 Feb 1997 23:09:31 -0800 (PST) Received: from root.com (implode.root.com [198.145.90.17]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id XAA02742; Mon, 17 Feb 1997 23:09:25 -0800 (PST) Received: from localhost (localhost [127.0.0.1]) by root.com (8.8.5/8.6.5) with SMTP id XAA04401; Mon, 17 Feb 1997 23:10:41 -0800 (PST) Message-Id: <199702180710.XAA04401@root.com> X-Authentication-Warning: implode.root.com: localhost [127.0.0.1] didn't use HELO protocol To: Bruce Evans cc: CVS-committers@freefall.freebsd.org, cvs-all@freefall.freebsd.org, cvs-sys@freefall.freebsd.org Subject: Re: cvs commit: src/sys/kern vfs_lookup.c In-reply-to: Your message of "Mon, 17 Feb 1997 22:46:43 PST." <199702180646.WAA00450@freefall.freebsd.org> From: David Greenman Reply-To: dg@root.com Date: Mon, 17 Feb 1997 23:10:41 -0800 Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >bde 97/02/17 22:46:42 > > Modified: sys/kern vfs_lookup.c > Log: > Fixed namei caching for LOOKUPs. It was broken for lstat() and olstat(). > Successful lstat()s purged an existing entry as well as not caching the > result. > > This bug was introduced in Lite1 by setting the LOCKPARENT flag for > [o]lstat() in order to support the inherit-attributes-from-parent- > directory misfeature for symlinks. LOCKPARENT was previously only set > for CREATEs and DELETEs. It is now set for LOOKUPs, but only for > [o]lstat(), so the problem wasn't very noticeable. We need to revert symlinks back to their traditional behavior (behave like they have an inode), too. I definately want to see this done for 3.0, and I wouldn't mind seeing it done prior to 2.2. NetBSD changed it back months ago. -DG David Greenman Core-team/Principal Architect, The FreeBSD Project From owner-cvs-sys Tue Feb 18 00:10:03 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id AAA05114 for cvs-sys-outgoing; Tue, 18 Feb 1997 00:10:03 -0800 (PST) Received: (from bde@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id AAA05067; Tue, 18 Feb 1997 00:09:56 -0800 (PST) Date: Tue, 18 Feb 1997 00:09:56 -0800 (PST) From: Bruce Evans Message-Id: <199702180809.AAA05067@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/sys resource.h Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk bde 97/02/18 00:09:53 Branch: sys/sys RELENG_2_2 Modified: sys/sys resource.h Log: YAMFC (quad_t -> rlim_t in struct rlimit). Revision Changes Path 1.4.4.1 +3 -3 src/sys/sys/resource.h From owner-cvs-sys Tue Feb 18 01:26:29 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id BAA19687 for cvs-sys-outgoing; Tue, 18 Feb 1997 01:26:29 -0800 (PST) Received: (from bde@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id BAA19667; Tue, 18 Feb 1997 01:26:25 -0800 (PST) Date: Tue, 18 Feb 1997 01:26:25 -0800 (PST) From: Bruce Evans Message-Id: <199702180926.BAA19667@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/sys tty.h Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk bde 97/02/18 01:26:23 Modified: sys/sys tty.h Log: Changed type of t_line from u_char to int. The Lite2 merge blew away rev.1.6 which changed it from char to int. Revision Changes Path 1.37 +1 -1 src/sys/sys/tty.h From owner-cvs-sys Tue Feb 18 06:07:18 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id GAA05388 for cvs-sys-outgoing; Tue, 18 Feb 1997 06:07:18 -0800 (PST) Received: (from bde@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id GAA05379; Tue, 18 Feb 1997 06:07:11 -0800 (PST) Date: Tue, 18 Feb 1997 06:07:11 -0800 (PST) From: Bruce Evans Message-Id: <199702181407.GAA05379@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/vm vm.h vm_map.h Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk bde 97/02/18 06:07:10 Modified: sys/vm vm.h vm_map.h Log: Removed vestiges of Mach lock types. vm_map.h: Removed #include of . curproc is only used in some macros and users of the macros already include . Revision Changes Path 1.12 +0 -11 src/sys/vm/vm.h 1.23 +1 -3 src/sys/vm/vm_map.h From owner-cvs-sys Tue Feb 18 06:37:30 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id GAA07105 for cvs-sys-outgoing; Tue, 18 Feb 1997 06:37:30 -0800 (PST) Received: (from bde@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id GAA07098; Tue, 18 Feb 1997 06:37:28 -0800 (PST) Date: Tue, 18 Feb 1997 06:37:28 -0800 (PST) From: Bruce Evans Message-Id: <199702181437.GAA07098@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/kern kern_lockf.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk bde 97/02/18 06:37:28 Modified: sys/kern kern_lockf.c Log: Added some ufs #includes so that this compiles with option LOCKF_DEBUG. Moving this all from ufs wasn't a good move. At least the debugging routines depend on the file system. Cleaned up the LOCKF_DEBUG #includes. Revision Changes Path 1.12 +7 -2 src/sys/kern/kern_lockf.c From owner-cvs-sys Tue Feb 18 07:03:23 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id HAA08375 for cvs-sys-outgoing; Tue, 18 Feb 1997 07:03:23 -0800 (PST) Received: (from bde@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id HAA08368; Tue, 18 Feb 1997 07:03:20 -0800 (PST) Date: Tue, 18 Feb 1997 07:03:20 -0800 (PST) From: Bruce Evans Message-Id: <199702181503.HAA08368@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/ufs/ufs quota.h Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk bde 97/02/18 07:03:20 Modified: sys/ufs/ufs quota.h Log: This now uses queue macros. Include if !KERNEL to preserve the documented interface. Revision Changes Path 1.9 +4 -0 src/sys/ufs/ufs/quota.h From owner-cvs-sys Tue Feb 18 07:09:44 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id HAA08757 for cvs-sys-outgoing; Tue, 18 Feb 1997 07:09:44 -0800 (PST) Received: (from bde@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id HAA08746; Tue, 18 Feb 1997 07:09:41 -0800 (PST) Date: Tue, 18 Feb 1997 07:09:41 -0800 (PST) From: Bruce Evans Message-Id: <199702181509.HAA08746@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/i386/boot/biosboot boot.h Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk bde 97/02/18 07:09:40 Modified: sys/i386/boot/biosboot boot.h Log: Fixed biosboot to compile with Lite2. is now a prerequisite for . Revision Changes Path 1.16 +1 -0 src/sys/i386/boot/biosboot/boot.h From owner-cvs-sys Tue Feb 18 08:06:21 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id IAA11787 for cvs-sys-outgoing; Tue, 18 Feb 1997 08:06:21 -0800 (PST) Received: from sovcom.kiae.su (sovcom.kiae.su [193.125.152.1]) by freefall.freebsd.org (8.8.5/8.8.5) with SMTP id IAA11766; Tue, 18 Feb 1997 08:06:01 -0800 (PST) Received: by sovcom.kiae.su id AA21003 (5.65.kiae-1 ); Tue, 18 Feb 1997 18:26:35 +0300 Received: by sovcom.KIAE.su (UUMAIL/2.0); Tue, 18 Feb 97 18:26:35 +0300 Received: (from ache@localhost) by nagual.ru (8.8.5/8.8.5) id SAA01611; Tue, 18 Feb 1997 18:19:23 +0300 (MSK) Date: Tue, 18 Feb 1997 18:19:19 +0300 (MSK) From: =?KOI8-R?B?4c7E0sXKIP7F0s7P1w==?= To: David Greenman Cc: Bruce Evans , CVS-committers@freefall.freebsd.org, cvs-all@freefall.freebsd.org, cvs-sys@freefall.freebsd.org Subject: Re: cvs commit: src/sys/kern vfs_lookup.c In-Reply-To: <199702180710.XAA04401@root.com> Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Mon, 17 Feb 1997, David Greenman wrote: > We need to revert symlinks back to their traditional behavior (behave like > they have an inode), too. I definately want to see this done for 3.0, and I > wouldn't mind seeing it done prior to 2.2. NetBSD changed it back months ago. I agree. New way looks like incomplete hack and cause too many complaints. -- Andrey A. Chernov http://www.nagual.ru/~ache/ From owner-cvs-sys Tue Feb 18 11:20:59 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id LAA23108 for cvs-sys-outgoing; Tue, 18 Feb 1997 11:20:59 -0800 (PST) Received: (from nate@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id LAA23095; Tue, 18 Feb 1997 11:20:57 -0800 (PST) Date: Tue, 18 Feb 1997 11:20:57 -0800 (PST) From: Nate Williams Message-Id: <199702181920.LAA23095@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/pccard pcic.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk nate 97/02/18 11:20:55 Modified: sys/pccard pcic.c Log: Try to make the BROKEN_VLSI chipset detection better, and not break non-broken chipssets whose ID is 0x84, such as the one found in the NEC 6030H. > The code relies on the assumption that on a genuine_broken vlsi, you > don't get 0x84 when probing slot 1 in the normal location. On the versa I > do get 0x84 when probing slot 1 in the normal location. What you get on > genuine_broken at the normal slot 1 location is unknown to me; Submitted by: Chris Timmons Revision Changes Path 1.31 +8 -2 src/sys/pccard/pcic.c From owner-cvs-sys Tue Feb 18 11:53:33 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id LAA26853 for cvs-sys-outgoing; Tue, 18 Feb 1997 11:53:33 -0800 (PST) Received: (from gibbs@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id LAA26846; Tue, 18 Feb 1997 11:53:30 -0800 (PST) Date: Tue, 18 Feb 1997 11:53:30 -0800 (PST) From: "Justin T. Gibbs" Message-Id: <199702181953.LAA26846@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/dev/aic7xxx aic7xxx.seq aic7xxx_reg.h src/sys/i386/scsi aic7xxx.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk gibbs 97/02/18 11:53:29 Branch: sys/dev/aic7xxx RELENG_2_1_0 Modified: sys/dev/aic7xxx aic7xxx.seq aic7xxx_reg.h Log: Enlarge the message out buffer from 6 to 8 bytes. Now that sync and wide negotiation messages may be tagged, we were overrunning the old buffer. The variable that was getting squashed is updated before the message goes out, causing corrupted SDTR or WDTR messages. Depending on the phases traversed before message out, this could cause the wrong offset to be negotiated allowing data overruns to occur. The problem is easier to detect with wide targets on the chain since the allowed offset is smaller. Also removed the unnecessary clearing of SPIORDY during the message out phase. We don't rely on SPIORDY any more. Revision Changes Path 1.16.4.20 +1 -2 src/sys/dev/aic7xxx/aic7xxx.seq 1.2.2.11 +12 -18 src/sys/dev/aic7xxx/aic7xxx_reg.h Branch: sys/i386/scsi RELENG_2_1_0 Modified: sys/i386/scsi aic7xxx.c Log: Kill the initialization of two old scratch ram variables. They were removed to make space for the larger message buffer. Revision Changes Path 1.29.2.30 +1 -13 src/sys/i386/scsi/aic7xxx.c From owner-cvs-sys Tue Feb 18 12:01:47 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id MAA27427 for cvs-sys-outgoing; Tue, 18 Feb 1997 12:01:47 -0800 (PST) Received: (from gibbs@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id MAA27417; Tue, 18 Feb 1997 12:01:42 -0800 (PST) Date: Tue, 18 Feb 1997 12:01:42 -0800 (PST) From: "Justin T. Gibbs" Message-Id: <199702182001.MAA27417@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/dev/aic7xxx aic7xxx.seq aic7xxx_reg.h src/sys/i386/scsi aic7xxx.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk gibbs 97/02/18 12:01:40 Branch: sys/dev/aic7xxx RELENG_2_2 Modified: sys/dev/aic7xxx aic7xxx.seq aic7xxx_reg.h Log: Enlarge the message out buffer from 6 to 8 bytes. Now that sync and wide negotiation messages may be tagged, we were overrunning the old buffer. The variable that was getting squashed is updated before the message goes out, causing corrupted SDTR or WDTR messages. Depending on the phases traversed before message out, this could cause the wrong offset to be negotiated allowing data overruns to occur. The problem is easier to detect with wide targets on the chain since the allowed offset is smaller. Also removed the unnecessary clearing of SPIORDY during the message out phase. We don't rely on SPIORDY any more. Revision Changes Path 1.46.2.6 +0 -1 src/sys/dev/aic7xxx/aic7xxx.seq 1.15.2.4 +11 -17 src/sys/dev/aic7xxx/aic7xxx_reg.h Branch: sys/i386/scsi RELENG_2_2 Modified: sys/i386/scsi aic7xxx.c Log: Kill the initialization of two old scratch ram variables. They were removed to make space for the larger message buffer. Revision Changes Path 1.81.2.10 +0 -12 src/sys/i386/scsi/aic7xxx.c From owner-cvs-sys Tue Feb 18 12:23:17 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id MAA28741 for cvs-sys-outgoing; Tue, 18 Feb 1997 12:23:17 -0800 (PST) Received: (from gibbs@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id MAA28733; Tue, 18 Feb 1997 12:23:13 -0800 (PST) Date: Tue, 18 Feb 1997 12:23:13 -0800 (PST) From: "Justin T. Gibbs" Message-Id: <199702182023.MAA28733@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/dev/aic7xxx aic7xxx.seq aic7xxx_reg.h src/sys/i386/scsi aic7xxx.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk gibbs 97/02/18 12:23:12 Modified: sys/dev/aic7xxx aic7xxx.seq aic7xxx_reg.h Log: Enlarge the message out buffer from 6 to 8 bytes. Now that sync and wide negotiation messages may be tagged, we were overrunning the old buffer. The variable that was getting squashed is updated before the message goes out, causing corrupted SDTR or WDTR messages. Depending on the phases traversed before message out, this could cause the wrong offset to be negotiated allowing data overruns to occur. The problem is easier to detect with wide targets on the chain since the allowed offset is smaller. Also removed the unnecessary clearing of SPIORDY during the message out phase. We don't rely on SPIORDY any more. Revision Changes Path 1.60 +0 -1 src/sys/dev/aic7xxx/aic7xxx.seq 1.24 +11 -17 src/sys/dev/aic7xxx/aic7xxx_reg.h Modified: sys/i386/scsi aic7xxx.c Log: Kill the initialization of two old scratch ram variables. They were removed to make space for the larger message buffer. Revision Changes Path 1.100 +0 -12 src/sys/i386/scsi/aic7xxx.c From owner-cvs-sys Tue Feb 18 12:47:34 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id MAA00595 for cvs-sys-outgoing; Tue, 18 Feb 1997 12:47:34 -0800 (PST) Received: (from wollman@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id MAA00580; Tue, 18 Feb 1997 12:47:29 -0800 (PST) Date: Tue, 18 Feb 1997 12:47:29 -0800 (PST) From: Garrett Wollman Message-Id: <199702182047.MAA00580@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/kern uipc_mbuf.c uipc_socket2.c src/sys/sys protosw.h src/sys/netinet in.c in_pcb.c in_pcb.h in_proto.c in_var.h ip_var.h raw_ip.c tcp_usrreq.c udp_usrreq.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk wollman 97/02/18 12:47:28 Modified: sys/kern uipc_mbuf.c uipc_socket2.c Log: uipc_mbuf.c: do a better job of counting how often we have to wait for memory, or are denied a cluster. uipc_socket2.c: define some generic ``operation-not-supported'' entry points for pr_usrreqs. Revision Changes Path 1.28 +16 -6 src/sys/kern/uipc_mbuf.c 1.20 +35 -0 src/sys/kern/uipc_socket2.c Modified: sys/netinet in.c in_pcb.c in_pcb.h in_proto.c in_var.h ip_var.h raw_ip.c tcp_usrreq.c udp_usrreq.c Log: Convert raw IP from mondo-switch-statement-from-Hell to pr_usrreqs. Collapse duplicates with udp_usrreq.c and tcp_usrreq.c (calling the generic routines in uipc_socket2.c and in_pcb.c). Calling sockaddr()_ or peeraddr() on a detached socket now traps, rather than harmlessly returning an error; this should never happen. Allow the raw IP buffer sizes to be controlled via sysctl. Revision Changes Path 1.31 +1 -1 src/sys/netinet/in.c 1.26 +19 -6 src/sys/netinet/in_pcb.c 1.17 +2 -2 src/sys/netinet/in_pcb.h 1.38 +20 -10 src/sys/netinet/in_proto.c 1.24 +1 -1 src/sys/netinet/in_var.h 1.30 +1 -2 src/sys/netinet/ip_var.h 1.42 +123 -151 src/sys/netinet/raw_ip.c 1.29 +6 -55 src/sys/netinet/tcp_usrreq.c 1.34 +6 -95 src/sys/netinet/udp_usrreq.c Modified: sys/sys protosw.h Log: Declare the new generic EOPNOTSUPP routines. Revision Changes Path 1.14 +5 -0 src/sys/sys/protosw.h From owner-cvs-sys Tue Feb 18 14:40:37 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id OAA06352 for cvs-sys-outgoing; Tue, 18 Feb 1997 14:40:37 -0800 (PST) Received: (from davidg@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id OAA06341; Tue, 18 Feb 1997 14:40:34 -0800 (PST) Date: Tue, 18 Feb 1997 14:40:34 -0800 (PST) From: David Greenman Message-Id: <199702182240.OAA06341@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/dev/ccd ccd.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk davidg 97/02/18 14:40:34 Branch: sys/dev/ccd RELENG_2_1_0 Modified: sys/dev/ccd ccd.c Log: Brought in fix from rev 1.18: fix for ccd when using bounce buffers (init b_bufsize). Revision Changes Path 1.10.2.4 +3 -1 src/sys/dev/ccd/ccd.c From owner-cvs-sys Tue Feb 18 15:31:58 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id PAA09038 for cvs-sys-outgoing; Tue, 18 Feb 1997 15:31:58 -0800 (PST) Received: (from msmith@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id PAA09031; Tue, 18 Feb 1997 15:31:56 -0800 (PST) Date: Tue, 18 Feb 1997 15:31:56 -0800 (PST) From: Michael Smith Message-Id: <199702182331.PAA09031@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/i386/isa wd.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk msmith 97/02/18 15:31:56 Modified: sys/i386/isa wd.c Log: Use the same blocks->size-in-MB conversion algorithm as the SCSI code to avoid overflowing an intermediate value for disks > 2^32 bytes large. Revision Changes Path 1.124 +1 -1 src/sys/i386/isa/wd.c From owner-cvs-sys Tue Feb 18 17:01:05 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id RAA16736 for cvs-sys-outgoing; Tue, 18 Feb 1997 17:01:05 -0800 (PST) Received: (from gibbs@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id RAA16716; Tue, 18 Feb 1997 17:01:01 -0800 (PST) Date: Tue, 18 Feb 1997 17:01:01 -0800 (PST) From: "Justin T. Gibbs" Message-Id: <199702190101.RAA16716@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/i386/scsi aic7xxx.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk gibbs 97/02/18 17:00:58 Modified: sys/i386/scsi aic7xxx.c Log: Fix some more problems in the recovery code. Cleanup of the disconnected list was broken in the SCB paging case (confusion of NULLand SCB_LIST_NULL) Implement a clean mechanism for determining that we have exited the timeout state and test for this in ahc_done instead of all over the place. Bring back the use of AAP (Auto Access Pause) I don't think it was the true cause of the bus hangs people were reporting. We want to reset the bus if we've been through an Abort action, not if we are a recovery SCB (one implies the other, but not vice-versa). Revision Changes Path 1.101 +37 -20 src/sys/i386/scsi/aic7xxx.c From owner-cvs-sys Tue Feb 18 17:01:52 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id RAA16842 for cvs-sys-outgoing; Tue, 18 Feb 1997 17:01:52 -0800 (PST) Received: (from gibbs@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id RAA16826; Tue, 18 Feb 1997 17:01:46 -0800 (PST) Date: Tue, 18 Feb 1997 17:01:46 -0800 (PST) From: "Justin T. Gibbs" Message-Id: <199702190101.RAA16826@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/i386/scsi aic7xxx.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk gibbs 97/02/18 17:01:45 Branch: sys/i386/scsi RELENG_2_2 Modified: sys/i386/scsi aic7xxx.c Log: Sync with current. Revision Changes Path 1.81.2.11 +37 -20 src/sys/i386/scsi/aic7xxx.c From owner-cvs-sys Tue Feb 18 17:02:30 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id RAA16934 for cvs-sys-outgoing; Tue, 18 Feb 1997 17:02:30 -0800 (PST) Received: (from gibbs@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id RAA16927; Tue, 18 Feb 1997 17:02:28 -0800 (PST) Date: Tue, 18 Feb 1997 17:02:28 -0800 (PST) From: "Justin T. Gibbs" Message-Id: <199702190102.RAA16927@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/i386/scsi aic7xxx.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk gibbs 97/02/18 17:02:28 Branch: sys/i386/scsi RELENG_2_1_0 Modified: sys/i386/scsi aic7xxx.c Log: Sync with current. Revision Changes Path 1.29.2.31 +38 -21 src/sys/i386/scsi/aic7xxx.c From owner-cvs-sys Tue Feb 18 19:13:52 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id TAA25405 for cvs-sys-outgoing; Tue, 18 Feb 1997 19:13:52 -0800 (PST) Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id TAA25336; Tue, 18 Feb 1997 19:12:14 -0800 (PST) Received: (from bde@localhost) by godzilla.zeta.org.au (8.8.3/8.6.9) id NAA01572; Wed, 19 Feb 1997 13:58:33 +1100 Date: Wed, 19 Feb 1997 13:58:33 +1100 From: Bruce Evans Message-Id: <199702190258.NAA01572@godzilla.zeta.org.au> To: cvs-all@freefall.freebsd.org, CVS-committers@freefall.freebsd.org, cvs-sys@freefall.freebsd.org, gibbs@freefall.freebsd.org Subject: Re: cvs commit: src/sys/dev/aic7xxx aic7xxx.seq aic7xxx_reg.h src/sys/i386/scsi aic7xxx.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > Branch: sys/dev/aic7xxx RELENG_2_1_0 > Modified: sys/dev/aic7xxx aic7xxx.seq aic7xxx_reg.h > Log: > Enlarge the message out buffer from 6 to 8 bytes. Now that sync and wide > negotiation messages may be tagged, we were overrunning the old buffer. > ... Please don't repeat long log messages in branches. Bruce From owner-cvs-sys Tue Feb 18 19:51:39 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id TAA28726 for cvs-sys-outgoing; Tue, 18 Feb 1997 19:51:39 -0800 (PST) Received: (from davidg@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id TAA28693; Tue, 18 Feb 1997 19:51:36 -0800 (PST) Date: Tue, 18 Feb 1997 19:51:36 -0800 (PST) From: David Greenman Message-Id: <199702190351.TAA28693@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/kern kern_exec.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk davidg 97/02/18 19:51:35 Modified: sys/kern kern_exec.c Log: Fix from PR #2757: execve() clears the P_SUGID process flag in execve() if the binary executed does not have suid or sgid permission bits set. This also happens when the effective uid is different from the real uid or the effective gid is different from the real gid. Under these circumstances, the process still has set id privileges and the P_SUGID flag should not be cleared. Submitted by: Tor Egge Revision Changes Path 1.51 +3 -1 src/sys/kern/kern_exec.c From owner-cvs-sys Tue Feb 18 19:53:39 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id TAA28961 for cvs-sys-outgoing; Tue, 18 Feb 1997 19:53:39 -0800 (PST) Received: (from davidg@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id TAA28954; Tue, 18 Feb 1997 19:53:37 -0800 (PST) Date: Tue, 18 Feb 1997 19:53:37 -0800 (PST) From: David Greenman Message-Id: <199702190353.TAA28954@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/kern kern_exec.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk davidg 97/02/18 19:53:37 Branch: sys/kern RELENG_2_2 Modified: sys/kern kern_exec.c Log: Brought in change from rev 1.51: preserve P_SUGID if {u,g}gid != r{u,g}id Note that this doesn't actually fix any known security holes, it just protects against sloppy programming. Revision Changes Path 1.47.2.2 +4 -2 src/sys/kern/kern_exec.c From owner-cvs-sys Tue Feb 18 22:05:57 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id WAA05458 for cvs-sys-outgoing; Tue, 18 Feb 1997 22:05:57 -0800 (PST) Received: (from asami@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id WAA05449; Tue, 18 Feb 1997 22:05:53 -0800 (PST) Date: Tue, 18 Feb 1997 22:05:53 -0800 (PST) From: Satoshi Asami Message-Id: <199702190605.WAA05449@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/conf newvers.sh Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk asami 97/02/18 22:05:50 Modified: sys/conf newvers.sh Log: Set RELDATE=300000. With the parallel version development, it has become impractical to distinguish versions using "real" release dates, so might as well make it correspond to real version number (-current is on the 3.0 branch) so at least the feature increments are guaranteed to be linear. Silently approved by: current list Revision Changes Path 1.31 +1 -1 src/sys/conf/newvers.sh From owner-cvs-sys Tue Feb 18 22:08:52 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id WAA05783 for cvs-sys-outgoing; Tue, 18 Feb 1997 22:08:52 -0800 (PST) Received: (from asami@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id WAA05776; Tue, 18 Feb 1997 22:08:49 -0800 (PST) Date: Tue, 18 Feb 1997 22:08:49 -0800 (PST) From: Satoshi Asami Message-Id: <199702190608.WAA05776@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/conf newvers.sh Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk asami 97/02/18 22:08:48 Branch: sys/conf RELENG_2_2 Modified: sys/conf newvers.sh Log: Merge from -current (well, sort of). Make RELDATE=220000 so that we can add new releases (2.1.8, 2.1.9, whatever) without crossing lines again. Silently approved by: current, core Revision Changes Path 1.26.2.7 +2 -2 src/sys/conf/newvers.sh From owner-cvs-sys Tue Feb 18 22:17:56 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id WAA06310 for cvs-sys-outgoing; Tue, 18 Feb 1997 22:17:56 -0800 (PST) Received: from dfw-ix8.ix.netcom.com (dfw-ix8.ix.netcom.com [206.214.98.8]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id WAA06302; Tue, 18 Feb 1997 22:17:53 -0800 (PST) Received: (from smap@localhost) by dfw-ix8.ix.netcom.com (8.8.4/8.8.4) id AAA23253; Wed, 19 Feb 1997 00:17:09 -0600 (CST) Received: from wck-ca5-04.ix.netcom.com(199.35.213.164) by dfw-ix8.ix.netcom.com via smap (V1.3) id sma023211; Wed Feb 19 00:16:46 1997 Received: (from asami@localhost) by silvia.HIP.Berkeley.EDU (8.8.5/8.6.9) id WAA25852; Tue, 18 Feb 1997 22:16:38 -0800 (PST) Date: Tue, 18 Feb 1997 22:16:38 -0800 (PST) Message-Id: <199702190616.WAA25852@silvia.HIP.Berkeley.EDU> To: CVS-committers@freefall.freebsd.org, cvs-all@freefall.freebsd.org, cvs-sys@freefall.freebsd.org In-reply-to: <199702190608.WAA05776@freefall.freebsd.org> (message from Satoshi Asami on Tue, 18 Feb 1997 22:08:49 -0800 (PST)) Subject: Re: cvs commit: src/sys/conf newvers.sh From: asami@vader.cs.berkeley.edu (Satoshi Asami) Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk * Branch: sys/conf RELENG_2_2 * Merge from -current (well, sort of). Make RELDATE=220000 so that we In case you are wondering, I'm about to fix all the ports that rely on this. Satoshi From owner-cvs-sys Wed Feb 19 05:06:10 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id FAA21541 for cvs-sys-outgoing; Wed, 19 Feb 1997 05:06:10 -0800 (PST) Received: (from kato@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id FAA21531; Wed, 19 Feb 1997 05:06:07 -0800 (PST) Date: Wed, 19 Feb 1997 05:06:07 -0800 (PST) From: KATO Takenori Message-Id: <199702191306.FAA21531@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/pc98/boot/biosboot boot.h Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk kato 97/02/19 05:06:06 Modified: sys/pc98/boot/biosboot boot.h Log: Synchronize with sys/i386/boot/biosboot/boot.h revision 1.16. Revision Changes Path 1.7 +1 -0 src/sys/pc98/boot/biosboot/boot.h From owner-cvs-sys Wed Feb 19 05:19:12 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id FAA21829 for cvs-sys-outgoing; Wed, 19 Feb 1997 05:19:12 -0800 (PST) Received: (from kato@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id FAA21820; Wed, 19 Feb 1997 05:19:09 -0800 (PST) Date: Wed, 19 Feb 1997 05:19:09 -0800 (PST) From: KATO Takenori Message-Id: <199702191319.FAA21820@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/pc98/pc98 wd.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk kato 97/02/19 05:19:09 Modified: sys/pc98/pc98 wd.c Log: Synchronize with sys/i386/isa/wd.c revision 1.124. Revision Changes Path 1.18 +1 -1 src/sys/pc98/pc98/wd.c From owner-cvs-sys Wed Feb 19 06:02:34 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id GAA24119 for cvs-sys-outgoing; Wed, 19 Feb 1997 06:02:34 -0800 (PST) Received: (from darrenr@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id GAA24112; Wed, 19 Feb 1997 06:02:32 -0800 (PST) Date: Wed, 19 Feb 1997 06:02:32 -0800 (PST) From: Darren Reed Message-Id: <199702191402.GAA24112@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/netinet ip_input.c ip_output.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk darrenr 97/02/19 06:02:31 Modified: sys/netinet ip_input.c ip_output.c Log: change IP Filter hooks to match new 3.1.8 patches for FreeBSD Revision Changes Path 1.59 +7 -6 src/sys/netinet/ip_input.c 1.50 +9 -9 src/sys/netinet/ip_output.c From owner-cvs-sys Wed Feb 19 06:30:44 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id GAA25413 for cvs-sys-outgoing; Wed, 19 Feb 1997 06:30:44 -0800 (PST) Received: (from kato@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id GAA25406; Wed, 19 Feb 1997 06:30:43 -0800 (PST) Date: Wed, 19 Feb 1997 06:30:43 -0800 (PST) From: KATO Takenori Message-Id: <199702191430.GAA25406@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/nfs nfs_nqlease.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk kato 97/02/19 06:30:42 Modified: sys/nfs nfs_nqlease.c Log: Moved nqnfs_vop_lease_check() inside #ifndef NFS_NOSERVER. So, NFS_NOSERVER kernel can be compiled again. Revision Changes Path 1.23 +2 -2 src/sys/nfs/nfs_nqlease.c From owner-cvs-sys Wed Feb 19 09:58:36 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id JAA07021 for cvs-sys-outgoing; Wed, 19 Feb 1997 09:58:36 -0800 (PST) Received: (from davidg@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id JAA07012; Wed, 19 Feb 1997 09:58:34 -0800 (PST) Date: Wed, 19 Feb 1997 09:58:34 -0800 (PST) From: David Greenman Message-Id: <199702191758.JAA07012@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/kern kern_exec.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk davidg 97/02/19 09:58:33 Branch: sys/kern RELENG_2_1_0 Modified: sys/kern kern_exec.c Log: Brought in fix from rev 1.51: don't clear P_SUGID if various {u,g}id don't match. Revision Changes Path 1.21.4.7 +4 -2 src/sys/kern/kern_exec.c From owner-cvs-sys Wed Feb 19 11:15:46 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id LAA12676 for cvs-sys-outgoing; Wed, 19 Feb 1997 11:15:46 -0800 (PST) Received: (from wollman@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id LAA12669; Wed, 19 Feb 1997 11:15:44 -0800 (PST) Date: Wed, 19 Feb 1997 11:15:44 -0800 (PST) From: Garrett Wollman Message-Id: <199702191915.LAA12669@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/kern uipc_socket2.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk wollman 97/02/19 11:15:44 Modified: sys/kern uipc_socket2.c Log: Make the operation of sonewconn1() a bit clearer by calling pru_attach() before putting the new connection on the connection queue. Revision Changes Path 1.21 +7 -12 src/sys/kern/uipc_socket2.c From owner-cvs-sys Thu Feb 20 03:51:56 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id DAA07907 for cvs-sys-outgoing; Thu, 20 Feb 1997 03:51:56 -0800 (PST) Received: (from bde@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id DAA07900; Thu, 20 Feb 1997 03:51:54 -0800 (PST) Date: Thu, 20 Feb 1997 03:51:54 -0800 (PST) From: Bruce Evans Message-Id: <199702201151.DAA07900@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/kern sys_generic.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk bde 97/02/20 03:51:53 Modified: sys/kern sys_generic.c Log: Improved select(): - avoid malloc() if the number of fds is small. - pack the bits better so that `small' is quite large. - don't waste time generating zero bits for null fd_set pointers or scanning these bits. Possibly improved select(): - free malloc()ed storage before returning. This is simpler and I think huge select()s aren't worth optimizing since they are rare, relative gain would be small and there would be tiny costs for all selects(). Reviewed by: ache (first version by him too) Revision Changes Path 1.22 +51 -36 src/sys/kern/sys_generic.c From owner-cvs-sys Thu Feb 20 05:30:43 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id FAA12358 for cvs-sys-outgoing; Thu, 20 Feb 1997 05:30:43 -0800 (PST) Received: from sovcom.kiae.su (sovcom.kiae.su [193.125.152.1]) by freefall.freebsd.org (8.8.5/8.8.5) with SMTP id FAA12342; Thu, 20 Feb 1997 05:30:32 -0800 (PST) Received: by sovcom.kiae.su id AA09480 (5.65.kiae-1 ); Thu, 20 Feb 1997 16:06:10 +0300 Received: by sovcom.KIAE.su (UUMAIL/2.0); Thu, 20 Feb 97 16:06:10 +0300 Received: (from ache@localhost) by nagual.ru (8.8.5/8.8.5) id QAA00747; Thu, 20 Feb 1997 16:03:57 +0300 (MSK) Date: Thu, 20 Feb 1997 16:03:52 +0300 (MSK) From: =?KOI8-R?B?4c7E0sXKIP7F0s7P1w==?= To: Bruce Evans Cc: CVS-committers@freefall.freebsd.org, cvs-all@freefall.freebsd.org, cvs-sys@freefall.freebsd.org Subject: Re: cvs commit: src/sys/kern sys_generic.c In-Reply-To: <199702201151.DAA07900@freefall.freebsd.org> Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Thu, 20 Feb 1997, Bruce Evans wrote: > bde 97/02/20 03:51:53 > > Modified: sys/kern sys_generic.c > Log: > Improved select(): > - avoid malloc() if the number of fds is small. > - pack the bits better so that `small' is quite large. > - don't waste time generating zero bits for null fd_set pointers or > scanning these bits. > > Possibly improved select(): > - free malloc()ed storage before returning. This is simpler and I > think huge select()s aren't worth optimizing since they are rare, > relative gain would be small and there would be tiny costs for all > selects(). Thanks, but I see no commit for removing p_selbits_size and p_selbits fields from struct proc... kern_exit.c also does free for p_selbits and needs fixing after remove. BTW, do we really need free() of malloc storage before return? If program use huge select once, in most practical cases it will mean that the same select code will be executed again, i.e. the same hudge select will be malloced/freeed. Maybe return to old p_selbits caching scheme? -- Andrey A. Chernov http://www.nagual.ru/~ache/ From owner-cvs-sys Thu Feb 20 05:53:34 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id FAA13445 for cvs-sys-outgoing; Thu, 20 Feb 1997 05:53:34 -0800 (PST) Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id FAA13400; Thu, 20 Feb 1997 05:52:57 -0800 (PST) Received: (from bde@localhost) by godzilla.zeta.org.au (8.8.3/8.6.9) id AAA02079; Fri, 21 Feb 1997 00:47:02 +1100 Date: Fri, 21 Feb 1997 00:47:02 +1100 From: Bruce Evans Message-Id: <199702201347.AAA02079@godzilla.zeta.org.au> To: ache@nagual.ru, bde@freefall.freebsd.org Subject: Re: cvs commit: src/sys/kern sys_generic.c Cc: cvs-all@freefall.freebsd.org, CVS-committers@freefall.freebsd.org, cvs-sys@freefall.freebsd.org Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >Thanks, but I see no commit for removing p_selbits_size and p_selbits >fields from struct proc... kern_exit.c also does free for p_selbits and >needs fixing after remove. I'll leave them until it is decided that the caching scheme is unnecssary. >BTW, do we really need free() of malloc storage before return? If program >use huge select once, in most practical cases it will mean that the same >select code will be executed again, i.e. the same hudge select will be >malloced/freeed. Maybe return to old p_selbits caching scheme? Maybe, but I think there are more important things to cache, starting with the 1K namei buffer. Most processes don't do many lookups, but things like `find / -xdev -type f' do a lot (about 4 per file). I think this isn't worth optimizing either :-). Bruce From owner-cvs-sys Thu Feb 20 10:39:37 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id KAA28634 for cvs-sys-outgoing; Thu, 20 Feb 1997 10:39:37 -0800 (PST) Received: (from wollman@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id KAA28622; Thu, 20 Feb 1997 10:39:35 -0800 (PST) Date: Thu, 20 Feb 1997 10:39:35 -0800 (PST) From: Garrett Wollman Message-Id: <199702201839.KAA28622@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/netinet ip_divert.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk wollman 97/02/20 10:39:34 Modified: sys/netinet ip_divert.c Log: Fix the parameters of a call to in_setsockaddr(). Revision Changes Path 1.5 +1 -1 src/sys/netinet/ip_divert.c From owner-cvs-sys Thu Feb 20 17:29:34 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id RAA24312 for cvs-sys-outgoing; Thu, 20 Feb 1997 17:29:34 -0800 (PST) Received: (from nate@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id RAA24305; Thu, 20 Feb 1997 17:29:33 -0800 (PST) Date: Thu, 20 Feb 1997 17:29:33 -0800 (PST) From: Nate Williams Message-Id: <199702210129.RAA24305@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/pccard pcic.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk nate 97/02/20 17:29:33 Modified: sys/pccard pcic.c Log: Whoops, make sure we have enough parenthesis. Pointed out by: bde Revision Changes Path 1.32 +1 -1 src/sys/pccard/pcic.c From owner-cvs-sys Thu Feb 20 21:47:06 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id VAA05760 for cvs-sys-outgoing; Thu, 20 Feb 1997 21:47:06 -0800 (PST) Received: (from nate@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id VAA05742; Thu, 20 Feb 1997 21:47:02 -0800 (PST) Date: Thu, 20 Feb 1997 21:47:02 -0800 (PST) From: Nate Williams Message-Id: <199702210547.VAA05742@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/pccard pcic.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk nate 97/02/20 21:47:01 Modified: sys/pccard pcic.c Log: Argh, this time get the parentheses right. This hasn't been a good day for me. Revision Changes Path 1.33 +3 -3 src/sys/pccard/pcic.c From owner-cvs-sys Fri Feb 21 08:30:35 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id IAA03560 for cvs-sys-outgoing; Fri, 21 Feb 1997 08:30:35 -0800 (PST) Received: (from wollman@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id IAA03553; Fri, 21 Feb 1997 08:30:33 -0800 (PST) Date: Fri, 21 Feb 1997 08:30:33 -0800 (PST) From: Garrett Wollman Message-Id: <199702211630.IAA03553@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/netinet tcp_usrreq.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk wollman 97/02/21 08:30:32 Modified: sys/netinet tcp_usrreq.c Log: Fix potential crash where a user attempts to perform an implied connect in TCP while sending urgent data. It is not clear what purpose is served by doing this, but there's no good reason why it shouldn't work. Submitted by: tjevans@raleigh.ibm.com via wpaul Revision Changes Path 1.30 +14 -1 src/sys/netinet/tcp_usrreq.c From owner-cvs-sys Fri Feb 21 08:48:40 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id IAA04813 for cvs-sys-outgoing; Fri, 21 Feb 1997 08:48:40 -0800 (PST) Received: (from wollman@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id IAA04805; Fri, 21 Feb 1997 08:48:37 -0800 (PST) Date: Fri, 21 Feb 1997 08:48:37 -0800 (PST) From: Garrett Wollman Message-Id: <199702211648.IAA04805@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/netinet tcp_usrreq.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk wollman 97/02/21 08:48:36 Branch: sys/netinet RELENG_2_2 Modified: sys/netinet tcp_usrreq.c Log: Daring greatly, since everyone else seems to defer to him in these matters, ourt fearless network person dons his asbestos suit and commits a security/crash bug fix to the 2.2 release branch. This merge includes revs. 1.28 (delete mondo switch statement form Hell) and 1.30 (fix bug where ordinary users could panic the system)). Rev. 1.29 was not merged because of outside dependencies. Revision Changes Path 1.26.2.1 +14 -318 src/sys/netinet/tcp_usrreq.c From owner-cvs-sys Fri Feb 21 09:05:59 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id JAA05756 for cvs-sys-outgoing; Fri, 21 Feb 1997 09:05:59 -0800 (PST) Received: (from wollman@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id JAA05745; Fri, 21 Feb 1997 09:05:55 -0800 (PST) Date: Fri, 21 Feb 1997 09:05:55 -0800 (PST) From: Garrett Wollman Message-Id: <199702211705.JAA05745@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/netinet tcp_usrreq.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk wollman 97/02/21 09:05:54 Branch: sys/netinet RELENG_2_1_0 Modified: sys/netinet tcp_usrreq.c Log: Make the analogous fix to rev. 1.30. Note that this TCP probably has lots of other bugs still remaining, and it also still has the mondo switch statement from Hell. Revision Changes Path 1.15.2.5 +14 -1 src/sys/netinet/tcp_usrreq.c From owner-cvs-sys Fri Feb 21 09:30:20 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id JAA07030 for cvs-sys-outgoing; Fri, 21 Feb 1997 09:30:20 -0800 (PST) Received: (from fenner@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id JAA07021; Fri, 21 Feb 1997 09:30:19 -0800 (PST) Date: Fri, 21 Feb 1997 09:30:19 -0800 (PST) From: Bill Fenner Message-Id: <199702211730.JAA07021@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/net rtsock.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk fenner 97/02/21 09:30:18 Branch: sys/net RELENG_2_2 Modified: sys/net rtsock.c Log: Bring in fix for PR#kern/2647 & kern/2783 from rev 1.25 . Ok'd by: joerg Revision Changes Path 1.20.2.1 +4 -4 src/sys/net/rtsock.c From owner-cvs-sys Fri Feb 21 09:35:40 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id JAA07569 for cvs-sys-outgoing; Fri, 21 Feb 1997 09:35:40 -0800 (PST) Received: (from fenner@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id JAA07562; Fri, 21 Feb 1997 09:35:39 -0800 (PST) Date: Fri, 21 Feb 1997 09:35:39 -0800 (PST) From: Bill Fenner Message-Id: <199702211735.JAA07562@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/net if_tun.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk fenner 97/02/21 09:35:38 Branch: sys/net RELENG_2_2 Modified: sys/net if_tun.c Log: Bring in fix for PR#misc/2733 from rev 1.32: use interface MTU instead of TUNMTU constant, and allow "ifconfig ... mtu" Ok'd by: jkh Revision Changes Path 1.29.2.2 +6 -1 src/sys/net/if_tun.c From owner-cvs-sys Fri Feb 21 10:35:09 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id KAA11382 for cvs-sys-outgoing; Fri, 21 Feb 1997 10:35:09 -0800 (PST) Received: (from wollman@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id KAA11374; Fri, 21 Feb 1997 10:35:06 -0800 (PST) Date: Fri, 21 Feb 1997 10:35:06 -0800 (PST) From: Garrett Wollman Message-Id: <199702211835.KAA11374@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/netinet ip_mroute.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk wollman 97/02/21 10:35:05 Modified: sys/netinet ip_mroute.c Log: Properly notice error returns from if_allmulti(). Revision Changes Path 1.38 +2 -2 src/sys/netinet/ip_mroute.c From owner-cvs-sys Fri Feb 21 15:15:24 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id PAA01982 for cvs-sys-outgoing; Fri, 21 Feb 1997 15:15:24 -0800 (PST) Received: (from ache@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id PAA01969; Fri, 21 Feb 1997 15:15:21 -0800 (PST) Date: Fri, 21 Feb 1997 15:15:21 -0800 (PST) From: "Andrey A. Chernov" Message-Id: <199702212315.PAA01969@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/sys types.h Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk ache 97/02/21 15:15:21 Modified: sys/sys types.h Log: Increase FD_SETSIZE back to 1024, select code fixed now Revision Changes Path 1.17 +1 -1 src/sys/sys/types.h From owner-cvs-sys Fri Feb 21 17:21:01 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id RAA10221 for cvs-sys-outgoing; Fri, 21 Feb 1997 17:21:01 -0800 (PST) Received: (from mpp@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id RAA10210; Fri, 21 Feb 1997 17:20:59 -0800 (PST) Date: Fri, 21 Feb 1997 17:20:59 -0800 (PST) From: Mike Pritchard Message-Id: <199702220120.RAA10210@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/kern kern_lkm.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk mpp 97/02/21 17:20:58 Modified: sys/kern kern_lkm.c Log: Support the installation of character devices via an lkm. Closes PR# 1716 Submitted by: jpt@magic.net Revision Changes Path 1.35 +10 -1 src/sys/kern/kern_lkm.c From owner-cvs-sys Sat Feb 22 04:52:41 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id EAA08894 for cvs-sys-outgoing; Sat, 22 Feb 1997 04:52:41 -0800 (PST) Received: (from peter@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id EAA08879; Sat, 22 Feb 1997 04:52:38 -0800 (PST) Date: Sat, 22 Feb 1997 04:52:38 -0800 (PST) From: Peter Wemm Message-Id: <199702221252.EAA08879@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/sys syscall-hide.h syscall.h sysproto.h src/sys/kern init_sysent.c syscalls.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk peter 97/02/22 04:52:37 Modified: sys/sys syscall-hide.h syscall.h sysproto.h sys/kern init_sysent.c syscalls.c Log: Regenerate to include correct Id string Revision Changes Path 1.28 +1 -1 src/sys/sys/syscall-hide.h 1.32 +1 -1 src/sys/sys/syscall.h 1.18 +1 -1 src/sys/sys/sysproto.h 1.39 +1 -1 src/sys/kern/init_sysent.c 1.34 +1 -1 src/sys/kern/syscalls.c From owner-cvs-sys Sat Feb 22 10:02:11 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id KAA29320 for cvs-sys-outgoing; Sat, 22 Feb 1997 10:02:11 -0800 (PST) Received: (from joerg@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id KAA29313; Sat, 22 Feb 1997 10:02:09 -0800 (PST) Date: Sat, 22 Feb 1997 10:02:09 -0800 (PST) From: Joerg Wunsch Message-Id: <199702221802.KAA29313@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/i386/isa wd.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk joerg 97/02/22 10:02:08 Branch: sys/i386/isa RELENG_2_2 Modified: sys/i386/isa wd.c Log: YAMFC (rev 1.124: prevent numerical overflow) Revision Changes Path 1.119.2.3 +2 -2 src/sys/i386/isa/wd.c From owner-cvs-sys Sat Feb 22 11:47:37 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id LAA10087 for cvs-sys-outgoing; Sat, 22 Feb 1997 11:47:37 -0800 (PST) Received: (from joerg@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id LAA10071; Sat, 22 Feb 1997 11:47:33 -0800 (PST) Date: Sat, 22 Feb 1997 11:47:33 -0800 (PST) From: Joerg Wunsch Message-Id: <199702221947.LAA10071@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/netinet ip_mroute.h Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk joerg 97/02/22 11:47:31 Branch: sys/netinet RELENG_2_2 Modified: sys/netinet ip_mroute.h Log: YAMFC (rev 1.11, fixes PR#2686) Revision Changes Path 1.10.4.1 +3 -3 src/sys/netinet/ip_mroute.h From owner-cvs-sys Sat Feb 22 12:31:31 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id MAA16111 for cvs-sys-outgoing; Sat, 22 Feb 1997 12:31:31 -0800 (PST) Received: (from joerg@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id MAA16103; Sat, 22 Feb 1997 12:31:29 -0800 (PST) Date: Sat, 22 Feb 1997 12:31:29 -0800 (PST) From: Joerg Wunsch Message-Id: <199702222031.MAA16103@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/i386/conf GENERIC LINT Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk joerg 97/02/22 12:31:28 Branch: sys/i386/conf RELENG_2_2 Modified: sys/i386/conf GENERIC LINT Log: Enable ex0. It has been commented out accidentally only. Revision Changes Path 1.77.2.4 +2 -2 src/sys/i386/conf/GENERIC 1.286.2.14 +2 -2 src/sys/i386/conf/LINT From owner-cvs-sys Sat Feb 22 12:50:46 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id MAA17693 for cvs-sys-outgoing; Sat, 22 Feb 1997 12:50:46 -0800 (PST) Received: (from joerg@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id MAA17685; Sat, 22 Feb 1997 12:50:44 -0800 (PST) Date: Sat, 22 Feb 1997 12:50:44 -0800 (PST) From: Joerg Wunsch Message-Id: <199702222050.MAA17685@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/kern kern_lkm.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk joerg 97/02/22 12:50:44 Branch: sys/kern RELENG_2_2 Modified: sys/kern kern_lkm.c Log: YAMFC (rev 1.35: allow for cdev LKMs) Revision Changes Path 1.32.2.1 +11 -2 src/sys/kern/kern_lkm.c From owner-cvs-sys Sat Feb 22 21:21:58 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id VAA18941 for cvs-sys-outgoing; Sat, 22 Feb 1997 21:21:58 -0800 (PST) Received: (from kato@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id VAA18934; Sat, 22 Feb 1997 21:21:56 -0800 (PST) Date: Sat, 22 Feb 1997 21:21:56 -0800 (PST) From: KATO Takenori Message-Id: <199702230521.VAA18934@freefall.freebsd.org> To: CVS-committers, cvs-all, cvs-sys Subject: cvs commit: src/sys/pc98/pc98 wd.c Sender: owner-cvs-sys@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk kato 97/02/22 21:21:56 Branch: sys/pc98/pc98 RELENG_2_2 Modified: sys/pc98/pc98 wd.c Log: Synchronize with sys/i386/isa/wd.c revision 1.119.2.3. Revision Changes Path 1.9.2.7 +2 -2 src/sys/pc98/pc98/wd.c