Re: svn commit: r343030 - in head/sys: cam conf dev/md dev/nvme fs/fuse fs/nfsclient fs/smbfs kern sys ufs/ffs vm
On Thu, 14 Feb 2019, Mark Johnston wrote: On Thu, Feb 14, 2019 at 06:56:42PM +1100, Bruce Evans wrote: * ... The only relevant commit between the good and bad versions seems to be r343453. This fixes uma_prealloc() to actually work. But it is a feature for it to not work when its caller asks for too much. I guess you meant r343353. In any case, the pbuf keg is _NOFREE, so even without preallocation the large pbuf zone limits may become problematic if there are bursts of allocation requests. Oops. * ... I don't understand how pbuf_preallocate() allocates for the other pbuf pools. When I debugged this for clpbufs, the preallocation was not used. pbuf types other than clpbufs seem to be unused in my configurations. I thought that pbufs were used during initialization, since they end up with a nonzero FREE count, but their only use seems to be to preallocate them. All of the pbuf zones share a common slab allocator. The zones have individual limits but can tap in to the shared preallocation. It seems to be working as intended now (except the allocation count is 3 higher than expected): XX ITEM SIZE LIMIT USED FREE REQ FAIL SLEEP XX XX swrbuf: 336,128, 0, 0, 0, 0, 0 XX swwbuf: 336, 64, 0, 0, 0, 0, 0 XX nfspbuf:336,128, 0, 0, 0, 0, 0 XX mdpbuf: 336, 25, 0, 0, 0, 0, 0 XX clpbuf: 336,128, 0, 35,2918, 0, 0 XX vnpbuf: 336, 2048, 0, 0, 0, 0, 0 XX pbuf: 336, 16, 0,2505, 0, 0, 0 pbuf should har 2537 preallocated and FREE initially, but seems to actually have 2540. pbufs were only used for clustering, and 35 of them were moved from pbuf to clpbuf. In the buggy version, the preallocations stopped after 4. Then clustering presumably moved these 4 to clpbuf. After that, clustering presumably used non-preallocated buffers until it reached its limit, and then recycled its own buffers. What should happen to recover the old overcommit behaviour with better debugging is 256 preallocated buffers (a few more for large systems) in pbuf and moving these to other pools, but never allocating from other pools (keep buffers in other pools only as an optimization and release them to the main pool under pressure). Also allow dynamic tuning of the pool[s] size[s]. The vnode cache does essentially this by using 1 overcommitted pool with unlimited size in uma and external management of the size. The separate pools correspond to separate file systems. These are too hard to manage, so the vnode cache throws everything into the main pool and depends on locality for the overcommit to not be too large. Bruce ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r344118 - head/sys/i386/include
On Thu, Feb 14, 2019 at 01:53:11PM +, Konstantin Belousov wrote: > New Revision: 344118 > URL: https://svnweb.freebsd.org/changeset/base/344118 > > Log: > Provide userspace versions of do_cpuid() and cpuid_count() on i386. > > Some older compilers, when generating PIC code, cannot handle inline > asm that clobbers %ebx (because %ebx is used as the GOT offset > register). Userspace versions avoid clobbering %ebx by saving it to > stack before executing the CPUID instruction. > > ... > +static __inline void > +do_cpuid(u_int ax, u_int *p) > +{ > + __asm __volatile( > + "pushl\t%%ebx\n\t" > + "cpuid\n\t" > + "movl\t%%ebx,%1\n\t" > + "popl\t%%ebx" Is there a reason to prefer pushl+movl+popl instead of movl+xchgl? "movl %%ebx, %1\n\t" "cpuid\n\t" "xchgl %%ebx, %1" ./danfe ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r343030 - in head/sys: cam conf dev/md dev/nvme fs/fuse fs/nfsclient fs/smbfs kern sys ufs/ffs vm
On Thu, 14 Feb 2019, Gleb Smirnoff wrote: On Wed, Feb 13, 2019 at 07:24:50PM -0600, Justin Hibbits wrote: J> This seems to break 32-bit platforms, or at least 32-bit book-e J> powerpc, which has a limited KVA space (~500MB). It preallocates I've J> seen over 2500 pbufs, at 128kB each, eating up over 300MB KVA, J> leaving very little left for the rest of runtime. J> J> I spent a couple hours earlier today debugging with Mark Johnston, and J> his consensus is that the vnode_pbuf_zone is too big on 32-bit J> platforms. Unfortunately I know very little about this area, so can't J> provide much extra insight, but can readily reproduce the issues I see J> triggered by this change, so am willing to help where I can. Ok, let's roll back to old default on 32-bit platforms and somewhat reduce the default on 64-bits. This reduces the largest allocation by a factor of 16 on 32-bit arches, (back to where it was), but it leves the other allocations unchanged, so the total allocation is still almost 5 times larger than before (down from 20 times larger). E.g., with the usual limit of 256 on nswbuf, the total allocation was 32MB with overcommit by a factor of about 5/2 on all systems, but it is now almost 80MB with no overcommit on 32-bit systems. Approximately 0MB of the extras are available on systems with 1GB kva, and less on systems with 512MB kva. Can you please confirm that the patch attached works for you? I don't have any systems affected by the bug, except when I boot with small hw.physmem or large kmem to test things. hw.physmem=72m leaves about 2MB afailable to map into buffers, and doesn't properly reduce nswbuf, so almost 80MB of kva is still used for pbufs. Allocating these must fail due to the RAM shortage. The old value of 32MB gives much the same failures (in practice, a larger operation like fork or exec tends to fail first). Limiting available kva is more interesting, and I haven't tested reducing it intentionally, except once I expanded kmem a lot to put a maximal md malloc()-backed disk in it). Expanding kmem steals from residual kva, and residual kva is not properly scaled except in my version. Large allocations then to cause panics at boot time, except for ones that crash because they don't check for errors. Here is debugging output for large allocations (1MB or more) at boot time on i386: XX pae_mode=0 with ~2.7 GB mapped RAM: XX kva_alloc: large allocation: 7490 pages: 0x580[0x1d42000] vm radix XX kva_alloc: large allocation: 6164 pages: 0x840[0x1814000] pmap init XX kva_alloc: large allocation: 28876 pages: 0xa00[0x70cc000] buf XX kmem_suballoc: large allocation: 1364 pages: 0x1140[0x554000] exec XX kmem_suballoc: large allocation: 10986 pages: 0x11954000[0x2aea000] pipe XX kva_alloc: large allocation: 6656 pages: 0x1480[0x1a0] sfbuf It went far above the old size of 1GB to nearly 1.5GB, but there is plenty to spare out of 4GB. Versions that fitted in 1GB started these allocations about 256MB lower and were otherwise similar. XX pae_mode=1 with 16 GB mapped RAM: XX kva_alloc: large allocation: 43832 pages: 0x14e0[0xab38000] vm radix XX kva_alloc: large allocation: 15668 pages: 0x2000[0x3d34000] pmap init XX kva_alloc: large allocation: 28876 pages: 0x23e0[0x70cc000] buf XX kmem_suballoc: large allocation: 1364 pages: 0x2b00[0x554000] exec XX kmem_suballoc: large allocation: 16320 pages: 0x2b554000[0x3fc] pipe XX kva_alloc: large allocation: 6656 pages: 0x2f60[0x1a0] sfbuf Only the vm radix and pmap init allocations are different, and they start much higher. The allocations now go over 3GB without any useful expansion except for the page tables. PAE was didn't work with 16 GB RAM and 1 GB kva, except in my version. PAE needed to be configured with 2 GB of kva to work with 16 GB RAM, but that was not the default or clearly documented. XX old PAE fixed fit work with 16GB RAM in 1GB KVA: XX kva_alloc: large allocation: 15691 pages: 0xd2c0[0x3d4b000] pmap init XX kva_alloc: large allocation: 43917 pages: 0xd6a0[0xab8d000] vm radix XX kva_alloc: large allocation: 27300 pages: 0xe160[0x6aa4000] buf XX kmem_suballoc: large allocation: 1364 pages: 0xe820[0x554000] exec XX kmem_suballoc: large allocation: 2291 pages: 0xe8754000[0x8f3000] pipe XX kva_alloc: large allocation: 6336 pages: 0xe920[0x18c]sfbuf PAE uses much more kva (almost 256MB extra) before the pmap and radix initializations here too. This is page table metadata before kva allocations are available. The fixes start by keeping track of this amout. It is about 1/16 of the address space for PAE in 1GB, so all later scaling was off by a factor of 16/15 (too high), and since there was less than 1/16 of 1GB to spare, PAE didn't fit. Only 'pipe' is reduced significantly to fit. swzone is reduced to 1 page in all cases, so it doesn't show here. It is about the same as sfbuf IIRC. The
svn commit: r344143 - head/sys/opencrypto
Author: sef Date: Fri Feb 15 04:15:43 2019 New Revision: 344143 URL: https://svnweb.freebsd.org/changeset/base/344143 Log: Fix another issue from r344141, having to do with size of a shift amount. This did not show up in my testing. Differential Revision:https://reviews.freebsd.org/D18592 Modified: head/sys/opencrypto/cbc_mac.c Modified: head/sys/opencrypto/cbc_mac.c == --- head/sys/opencrypto/cbc_mac.c Fri Feb 15 04:01:59 2019 (r344142) +++ head/sys/opencrypto/cbc_mac.c Fri Feb 15 04:15:43 2019 (r344143) @@ -128,7 +128,7 @@ AES_CBC_MAC_Reinit(struct aes_cbc_mac_ctx *ctx, const uint16_t sizeVal = htobe16(ctx->authDataLength); bcopy(, ctx->staging_block, sizeof(sizeVal)); ctx->blockIndex = sizeof(sizeVal); - } else if (ctx->authDataLength < (1UL<<32)) { + } else if (ctx->authDataLength < (1ULL<<32)) { uint32_t sizeVal = htobe32(ctx->authDataLength); ctx->staging_block[0] = 0xff; ctx->staging_block[1] = 0xfe; ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r344142 - head/sys/opencrypto
Author: sef Date: Fri Feb 15 04:01:59 2019 New Revision: 344142 URL: https://svnweb.freebsd.org/changeset/base/344142 Log: Pasting in a source control line missed the last quote. Fixed. Modified: head/sys/opencrypto/cbc_mac.c Modified: head/sys/opencrypto/cbc_mac.c == --- head/sys/opencrypto/cbc_mac.c Fri Feb 15 03:53:03 2019 (r344141) +++ head/sys/opencrypto/cbc_mac.c Fri Feb 15 04:01:59 2019 (r344142) @@ -23,7 +23,7 @@ */ #include -__FBSDID("$FreeBSD$); +__FBSDID("$FreeBSD$"); #include #include ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r344141 - in head: sys/opencrypto tools/tools/crypto
Author: sef Date: Fri Feb 15 03:53:03 2019 New Revision: 344141 URL: https://svnweb.freebsd.org/changeset/base/344141 Log: Add AES-CCM encryption, and plumb into OCF. This commit essentially has three parts: * Add the AES-CCM encryption hooks. This is in and of itself fairly small, as there is only a small difference between CCM and the other ICM-based algorithms. * Hook the code into the OpenCrypto framework. This is the bulk of the changes, as the algorithm type has to be checked for, and the differences between it and GCM dealt with. * Update the cryptocheck tool to be aware of it. This is invaluable for confirming that the code works. This is a software-only implementation, meaning that the performance is very low. Sponsored by: iXsystems Inc. Differential Revision:https://reviews.freebsd.org/D19090 Modified: head/sys/opencrypto/cryptodev.c head/sys/opencrypto/cryptodev.h head/sys/opencrypto/cryptosoft.c head/sys/opencrypto/xform_aes_icm.c head/sys/opencrypto/xform_auth.h head/sys/opencrypto/xform_enc.h head/tools/tools/crypto/cryptocheck.c Modified: head/sys/opencrypto/cryptodev.c == --- head/sys/opencrypto/cryptodev.c Fri Feb 15 03:46:39 2019 (r344140) +++ head/sys/opencrypto/cryptodev.c Fri Feb 15 03:53:03 2019 (r344141) @@ -444,6 +444,9 @@ cryptof_ioctl( case CRYPTO_CHACHA20: txform = _xform_chacha20; break; + case CRYPTO_AES_CCM_16: + txform = _xform_ccm; + break; default: CRYPTDEB("invalid cipher"); @@ -488,6 +491,25 @@ cryptof_ioctl( thash = _hash_nist_gmac_aes_256; break; + case CRYPTO_AES_CCM_CBC_MAC: + switch (sop->keylen) { + case 16: + thash = _hash_ccm_cbc_mac_128; + break; + case 24: + thash = _hash_ccm_cbc_mac_192; + break; + case 32: + thash = _hash_ccm_cbc_mac_256; + break; + default: + CRYPTDEB("Invalid CBC MAC key size %d", + sop->keylen); + SDT_PROBE1(opencrypto, dev, ioctl, + error, __LINE__); + return (EINVAL); + } + break; #ifdef notdef case CRYPTO_MD5: thash = _hash_md5; @@ -1003,12 +1025,13 @@ cryptodev_aead( } /* -* For GCM, crd_len covers only the AAD. For other ciphers +* For GCM/CCM, crd_len covers only the AAD. For other ciphers * chained with an HMAC, crd_len covers both the AAD and the * cipher text. */ crda->crd_skip = 0; - if (cse->cipher == CRYPTO_AES_NIST_GCM_16) + if (cse->cipher == CRYPTO_AES_NIST_GCM_16 || + cse->cipher == CRYPTO_AES_CCM_16) crda->crd_len = caead->aadlen; else crda->crd_len = caead->aadlen + caead->len; Modified: head/sys/opencrypto/cryptodev.h == --- head/sys/opencrypto/cryptodev.h Fri Feb 15 03:46:39 2019 (r344140) +++ head/sys/opencrypto/cryptodev.h Fri Feb 15 03:53:03 2019 (r344141) @@ -133,6 +133,7 @@ #defineARC4_IV_LEN 1 #defineAES_GCM_IV_LEN 12 +#defineAES_CCM_IV_LEN 12 #defineAES_XTS_IV_LEN 8 #defineAES_XTS_ALPHA 0x87/* GF(2^128) generator polynomial */ @@ -204,7 +205,8 @@ #defineCRYPTO_SHA2_512 37 #defineCRYPTO_POLY1305 38 #defineCRYPTO_AES_CCM_CBC_MAC 39 /* auth side */ -#defineCRYPTO_ALGORITHM_MAX39 /* Keep updated - see below */ +#defineCRYPTO_AES_CCM_16 40 /* cipher side */ +#defineCRYPTO_ALGORITHM_MAX40 /* Keep updated - see below */ #defineCRYPTO_ALGO_VALID(x)((x) >= CRYPTO_ALGORITHM_MIN && \ (x) <= CRYPTO_ALGORITHM_MAX) Modified: head/sys/opencrypto/cryptosoft.c == --- head/sys/opencrypto/cryptosoft.cFri Feb 15 03:46:39 2019 (r344140) +++ head/sys/opencrypto/cryptosoft.cFri Feb 15 03:53:03 2019 (r344141) @@ -62,6 +62,9 @@ __FBSDID("$FreeBSD$"); #include #include "cryptodev_if.h" +_Static_assert(AES_CCM_IV_LEN == AES_GCM_IV_LEN, +
svn commit: r344140 - in head/sys: conf modules/crypto opencrypto
Author: sef Date: Fri Feb 15 03:46:39 2019 New Revision: 344140 URL: https://svnweb.freebsd.org/changeset/base/344140 Log: Add CBC-MAC authentication. This adds the CBC-MAC code to the kernel, but does not hook it up to anything (that comes in the next commit). https://tools.ietf.org/html/rfc3610 describes the algorithm. Note that this is a software-only implementation, which means it is fairly slow. Sponsored by: iXsystems Inc Differential Revision: https://reviews.freebsd.org/D18592 Added: head/sys/opencrypto/cbc_mac.c (contents, props changed) head/sys/opencrypto/cbc_mac.h (contents, props changed) head/sys/opencrypto/xform_cbc_mac.c (contents, props changed) Modified: head/sys/conf/files head/sys/modules/crypto/Makefile head/sys/opencrypto/cryptodev.h Modified: head/sys/conf/files == --- head/sys/conf/files Fri Feb 15 00:29:44 2019(r344139) +++ head/sys/conf/files Fri Feb 15 03:46:39 2019(r344140) @@ -4847,6 +4847,8 @@ crypto/libsodium/randombytes.coptional crypto \ compile-with "${NORMAL_C} -I$S/contrib/libsodium/src/libsodium/include -I$S/crypto/libsodium" crypto/libsodium/utils.c optional crypto \ compile-with "${NORMAL_C} -I$S/contrib/libsodium/src/libsodium/include -I$S/crypto/libsodium" +opencrypto/cbc_mac.c optional crypto +opencrypto/xform_cbc_mac.c optional crypto rpc/auth_none.coptional krpc | nfslockd | nfscl | nfsd rpc/auth_unix.coptional krpc | nfslockd | nfscl | nfsd rpc/authunix_prot.coptional krpc | nfslockd | nfscl | nfsd Modified: head/sys/modules/crypto/Makefile == --- head/sys/modules/crypto/MakefileFri Feb 15 00:29:44 2019 (r344139) +++ head/sys/modules/crypto/MakefileFri Feb 15 03:46:39 2019 (r344140) @@ -68,5 +68,7 @@ CFLAGS.utils.c+= -I${LIBSODIUM_INC} -I${LIBSODIUM_C SRCS += opt_param.h cryptodev_if.h bus_if.h device_if.h SRCS += opt_ddb.h +SRCS += cbc_mac.c +SRCS += xform_cbc_mac.c .include Added: head/sys/opencrypto/cbc_mac.c == --- /dev/null 00:00:00 1970 (empty, because file is newly added) +++ head/sys/opencrypto/cbc_mac.c Fri Feb 15 03:46:39 2019 (r344140) @@ -0,0 +1,252 @@ +/* + * Copyright (c) 2018-2019 iXsystems Inc. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + *notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + *notice, this list of conditions and the following disclaimer in the + *documentation and/or other materials provided with the distribution. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR + * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES + * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. + * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, + * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT + * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF + * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ + +#include +__FBSDID("$FreeBSD$); + +#include +#include +#include +#include +#include +#include + +/* + * Given two CCM_CBC_BLOCK_LEN blocks, xor + * them into dst, and then encrypt dst. + */ +static void +xor_and_encrypt(struct aes_cbc_mac_ctx *ctx, + const uint8_t *src, uint8_t *dst) +{ + const uint64_t *b1; + uint64_t *b2; + uint64_t temp_block[CCM_CBC_BLOCK_LEN/sizeof(uint64_t)]; + + b1 = (const uint64_t*)src; + b2 = (uint64_t*)dst; + + for (size_t count = 0; +count < CCM_CBC_BLOCK_LEN/sizeof(uint64_t); +count++) { + temp_block[count] = b1[count] ^ b2[count]; + } + rijndaelEncrypt(ctx->keysched, ctx->rounds, (void*)temp_block, dst); +} + +void +AES_CBC_MAC_Init(struct aes_cbc_mac_ctx *ctx) +{ + bzero(ctx, sizeof(*ctx)); +} + +void +AES_CBC_MAC_Setkey(struct aes_cbc_mac_ctx *ctx, const uint8_t *key, uint16_t klen) +{ + ctx->rounds = rijndaelKeySetupEnc(ctx->keysched, key, klen * 8); +} + +/* + * This is called to set the nonce, aka IV. + * Before this call, the authDataLength and
Re: svn commit: r344129 - head
On Thu, Feb 14, 2019, 1:42 PM Rodney W. Grimes < free...@pdx.rh.cn85.dnsmgr.net wrote: > > On Thu, Feb 14, 2019 at 12:00:22PM -0700, Warner Losh wrote: > > > On Thu, Feb 14, 2019 at 11:29 AM Rodney W. Grimes < > > > > > Differential Review: https://reviews.freebsd.org/D19193 > > > > > > > > You sited a differential, but not give any attribution > > > > to the external source :-( > > > > > > > > > > The differential review has that information. > > > > External contributors should be recognized by having their names appear > > in the commit logs. > > We even bother to put a special line in the commit template > for this. Further it has been standard operating procedure > for at least as long as I have been back that submitters > are infact recognized in commit messages. > > You have, again, summarily dismissed valid feedback. > The problem is that was a trivial commit. And you offered not one, but two complaints about adding an 'e' to the updating file. That's what pissed me off. It's advice that might be correct, but was so far over the top, in public, for such a trivial commit. That's why I got mad: it added no value and seemed nit picky and pretty. So I lost it. I shouldn't have, but I did. I'm sorry for my cool. Warner > ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r343030 - in head/sys: cam conf dev/md dev/nvme fs/fuse fs/nfsclient fs/smbfs kern sys ufs/ffs vm
> On Feb 14, 2019, at 15:34, Gleb Smirnoff wrote: > > Hi Justin, > > On Wed, Feb 13, 2019 at 07:24:50PM -0600, Justin Hibbits wrote: > J> This seems to break 32-bit platforms, or at least 32-bit book-e > J> powerpc, which has a limited KVA space (~500MB). It preallocates I've > J> seen over 2500 pbufs, at 128kB each, eating up over 300MB KVA, > J> leaving very little left for the rest of runtime. > J> > J> I spent a couple hours earlier today debugging with Mark Johnston, and > J> his consensus is that the vnode_pbuf_zone is too big on 32-bit > J> platforms. Unfortunately I know very little about this area, so can't > J> provide much extra insight, but can readily reproduce the issues I see > J> triggered by this change, so am willing to help where I can. > > Ok, let's roll back to old default on 32-bit platforms and somewhat > reduce the default on 64-bits. > > Can you please confirm that the patch attached works for you? Quick question: why was the value reduced by a factor of 4 on 64-bit platforms? Thanks so much! -Enji ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r344139 - in stable/12: share/man/man4 sys/netgraph
Author: glebius Date: Fri Feb 15 00:29:44 2019 New Revision: 344139 URL: https://svnweb.freebsd.org/changeset/base/344139 Log: Merge r343895: Allow some nesting of ng_iface(4) interfaces and add a configuration knob. PR: 235500 Modified: stable/12/share/man/man4/ng_iface.4 stable/12/sys/netgraph/ng_iface.c Directory Properties: stable/12/ (props changed) Modified: stable/12/share/man/man4/ng_iface.4 == --- stable/12/share/man/man4/ng_iface.4 Thu Feb 14 20:33:35 2019 (r344138) +++ stable/12/share/man/man4/ng_iface.4 Fri Feb 15 00:29:44 2019 (r344139) @@ -35,7 +35,7 @@ .\" $FreeBSD$ .\" $Whistle: ng_iface.8,v 1.5 1999/01/25 23:46:26 archie Exp $ .\" -.Dd January 12, 2015 +.Dd February 6, 2019 .Dt NG_IFACE 4 .Os .Sh NAME @@ -144,6 +144,17 @@ In case when your graph ends up with some kind of seri synchronous or modem, the .Nm is the right place to turn ALTQ on. +.Sh Nesting +.Nm +supports nesting, a configuration when traffic of one +.Nm +interface flows through the other. +The default maximum allowed nesting level is 2. +It can be changed at runtime setting +.Xr sysctl 8 +variable +.Va net.graph.iface.max_nesting +to the desired level of nesting. .Sh SEE ALSO .Xr altq 4 , .Xr bpf 4 , @@ -151,6 +162,7 @@ is the right place to turn ALTQ on. .Xr ng_cisco 4 , .Xr ifconfig 8 , .Xr ngctl 8 +.Xr sysctl .Sh HISTORY The .Nm iface Modified: stable/12/sys/netgraph/ng_iface.c == --- stable/12/sys/netgraph/ng_iface.c Thu Feb 14 20:33:35 2019 (r344138) +++ stable/12/sys/netgraph/ng_iface.c Fri Feb 15 00:29:44 2019 (r344139) @@ -68,6 +68,7 @@ #include #include #include +#include #include #include @@ -92,6 +93,13 @@ static MALLOC_DEFINE(M_NETGRAPH_IFACE, "netgraph_iface #define M_NETGRAPH_IFACE M_NETGRAPH #endif +static SYSCTL_NODE(_net_graph, OID_AUTO, iface, CTLFLAG_RW, 0, +"Point to point netgraph interface"); +VNET_DEFINE_STATIC(int, ng_iface_max_nest) = 2; +#defineV_ng_iface_max_nest VNET(ng_iface_max_nest) +SYSCTL_INT(_net_graph_iface, OID_AUTO, max_nesting, CTLFLAG_VNET | CTLFLAG_RW, +_NAME(ng_iface_max_nest), 0, "Max nested tunnels"); + /* This struct describes one address family */ struct iffam { sa_family_t family; /* Address family */ @@ -355,7 +363,8 @@ ng_iface_output(struct ifnet *ifp, struct mbuf *m, } /* Protect from deadly infinite recursion. */ - error = if_tunnel_check_nesting(ifp, m, NGM_IFACE_COOKIE, 1); + error = if_tunnel_check_nesting(ifp, m, NGM_IFACE_COOKIE, + V_ng_iface_max_nest); if (error) { m_freem(m); return (error); ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r343030 - in head/sys: cam conf dev/md dev/nvme fs/fuse fs/nfsclient fs/smbfs kern sys ufs/ffs vm
Hi Justin, On Wed, Feb 13, 2019 at 07:24:50PM -0600, Justin Hibbits wrote: J> This seems to break 32-bit platforms, or at least 32-bit book-e J> powerpc, which has a limited KVA space (~500MB). It preallocates I've J> seen over 2500 pbufs, at 128kB each, eating up over 300MB KVA, J> leaving very little left for the rest of runtime. J> J> I spent a couple hours earlier today debugging with Mark Johnston, and J> his consensus is that the vnode_pbuf_zone is too big on 32-bit J> platforms. Unfortunately I know very little about this area, so can't J> provide much extra insight, but can readily reproduce the issues I see J> triggered by this change, so am willing to help where I can. Ok, let's roll back to old default on 32-bit platforms and somewhat reduce the default on 64-bits. Can you please confirm that the patch attached works for you? -- Gleb Smirnoff diff --git a/sys/vm/vnode_pager.c b/sys/vm/vnode_pager.c index 3e71ab4436cc..ded9e65e4e4c 100644 --- a/sys/vm/vnode_pager.c +++ b/sys/vm/vnode_pager.c @@ -115,13 +115,23 @@ SYSCTL_PROC(_debug, OID_AUTO, vnode_domainset, CTLTYPE_STRING | CTLFLAG_RW, _domainset, 0, sysctl_handle_domainset, "A", "Default vnode NUMA policy"); +static int nvnpbufs; +SYSCTL_INT(_vm, OID_AUTO, vnode_pbufs, CTLFLAG_RDTUN | CTLFLAG_NOFETCH, +, 0, "number of physical buffers allocated for vnode pager"); + static uma_zone_t vnode_pbuf_zone; static void vnode_pager_init(void *dummy) { - vnode_pbuf_zone = pbuf_zsecond_create("vnpbuf", nswbuf * 8); +#ifdef __LP64__ + nvnpbufs = nswbuf * 2; +#else + nvnpbufs = nswbuf / 2; +#endif + TUNABLE_INT_FETCH("vm.vnode_pbufs", ); + vnode_pbuf_zone = pbuf_zsecond_create("vnpbuf", nvnpbufs); } SYSINIT(vnode_pager, SI_SUB_CPU, SI_ORDER_ANY, vnode_pager_init, NULL); ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r344129 - head
On Thu, Feb 14, 2019, 1:48 PM Juli Mallett On Thu, 14 Feb 2019 at 12:42, Rodney W. Grimes < > free...@pdx.rh.cn85.dnsmgr.net> wrote: > >> > On Thu, Feb 14, 2019 at 12:00:22PM -0700, Warner Losh wrote: >> > > On Thu, Feb 14, 2019 at 11:29 AM Rodney W. Grimes < >> > > > > Differential Review: https://reviews.freebsd.org/D19193 >> > > > >> > > > You sited a differential, but not give any attribution >> > > > to the external source :-( >> > > > >> > > >> > > The differential review has that information. >> > >> > External contributors should be recognized by having their names appear >> > in the commit logs. >> >> We even bother to put a special line in the commit template >> for this. Further it has been standard operating procedure >> for at least as long as I have been back that submitters >> are infact recognized in commit messages. >> > > Yeah, Phabricator must not become a second source of truth for who > actually did the work, which we had at least one case of recently. That's > no good. In this case, the difference between a patch and a bug report is > indistinguishable ("word X is misspelt at offset Y in file Z" is probably > not really Phabricator fodder at the scale of a single instance), but it > should plainly either be a Reported byline if not a Submitted one in the > commit message itself. > This is the most pointless waste of time of the year. It was a one letter typo. Yes, maybe I should have included the submitter's name, but I didn't. It doesn't deserve a long thread and a snarky response. Geeze people, get some perspective. Warner > ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r344129 - head
On Thu, 14 Feb 2019 at 12:42, Rodney W. Grimes < free...@pdx.rh.cn85.dnsmgr.net> wrote: > > On Thu, Feb 14, 2019 at 12:00:22PM -0700, Warner Losh wrote: > > > On Thu, Feb 14, 2019 at 11:29 AM Rodney W. Grimes < > > > > > Differential Review: https://reviews.freebsd.org/D19193 > > > > > > > > You sited a differential, but not give any attribution > > > > to the external source :-( > > > > > > > > > > The differential review has that information. > > > > External contributors should be recognized by having their names appear > > in the commit logs. > > We even bother to put a special line in the commit template > for this. Further it has been standard operating procedure > for at least as long as I have been back that submitters > are infact recognized in commit messages. > Yeah, Phabricator must not become a second source of truth for who actually did the work, which we had at least one case of recently. That's no good. In this case, the difference between a patch and a bug report is indistinguishable ("word X is misspelt at offset Y in file Z" is probably not really Phabricator fodder at the scale of a single instance), but it should plainly either be a Reported byline if not a Submitted one in the commit message itself. Juli. ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r344129 - head
> On Thu, Feb 14, 2019 at 12:00:22PM -0700, Warner Losh wrote: > > On Thu, Feb 14, 2019 at 11:29 AM Rodney W. Grimes < > > > > Differential Review: https://reviews.freebsd.org/D19193 > > > > > > You sited a differential, but not give any attribution > > > to the external source :-( > > > > > > > The differential review has that information. > > External contributors should be recognized by having their names appear > in the commit logs. We even bother to put a special line in the commit template for this. Further it has been standard operating procedure for at least as long as I have been back that submitters are infact recognized in commit messages. You have, again, summarily dismissed valid feedback. -- Rod Grimes rgri...@freebsd.org ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r344138 - stable/12/sbin/ifconfig
Author: asomers Date: Thu Feb 14 20:33:35 2019 New Revision: 344138 URL: https://svnweb.freebsd.org/changeset/base/344138 Log: MFC r343530: ifconfig: fix endianness bug displaying pfsync interfaces Reviewed by: kp Differential Revision:https://reviews.freebsd.org/D19005 Modified: stable/12/sbin/ifconfig/ifpfsync.c Directory Properties: stable/12/ (props changed) Modified: stable/12/sbin/ifconfig/ifpfsync.c == --- stable/12/sbin/ifconfig/ifpfsync.c Thu Feb 14 20:31:05 2019 (r344137) +++ stable/12/sbin/ifconfig/ifpfsync.c Thu Feb 14 20:33:35 2019 (r344138) @@ -195,16 +195,16 @@ pfsync_status(int s) return; if (preq.pfsyncr_syncdev[0] != '\0' || - preq.pfsyncr_syncpeer.s_addr != INADDR_PFSYNC_GROUP) + preq.pfsyncr_syncpeer.s_addr != htonl(INADDR_PFSYNC_GROUP)) printf("\t"); if (preq.pfsyncr_syncdev[0] != '\0') printf("pfsync: syncdev: %s ", preq.pfsyncr_syncdev); - if (preq.pfsyncr_syncpeer.s_addr != INADDR_PFSYNC_GROUP) + if (preq.pfsyncr_syncpeer.s_addr != htonl(INADDR_PFSYNC_GROUP)) printf("syncpeer: %s ", inet_ntoa(preq.pfsyncr_syncpeer)); if (preq.pfsyncr_syncdev[0] != '\0' || - preq.pfsyncr_syncpeer.s_addr != INADDR_PFSYNC_GROUP) { + preq.pfsyncr_syncpeer.s_addr != htonl(INADDR_PFSYNC_GROUP)) { printf("maxupd: %d ", preq.pfsyncr_maxupdates); printf("defer: %s\n", preq.pfsyncr_defer ? "on" : "off"); } ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r344137 - stable/12/tools/build/mk
Author: asomers Date: Thu Feb 14 20:31:05 2019 New Revision: 344137 URL: https://svnweb.freebsd.org/changeset/base/344137 Log: MFC r300938, r342154 r300938 by ngie: Remove the sa(8) tests if MK_ACCT == no when "make delete-old" is run sa(8) is conditionally installed based on MK_ACCT != no today Sponsored by: EMC / Isilon Storage Division r342154: OptionalObsoleteFiles: Fix deleting usr/tests/usr.sbin/sa It's a directory, not a file. Reported by: ngie X-MFC-With: 300938 Modified: stable/12/tools/build/mk/OptionalObsoleteFiles.inc Directory Properties: stable/12/ (props changed) Modified: stable/12/tools/build/mk/OptionalObsoleteFiles.inc == --- stable/12/tools/build/mk/OptionalObsoleteFiles.inc Thu Feb 14 20:29:33 2019(r344136) +++ stable/12/tools/build/mk/OptionalObsoleteFiles.inc Thu Feb 14 20:31:05 2019(r344137) @@ -38,7 +38,7 @@ OLD_FILES+=usr/tests/usr.sbin/sa/v2-i386-usr.in OLD_FILES+=usr/tests/usr.sbin/sa/v2-sparc64-sav.in OLD_FILES+=usr/tests/usr.sbin/sa/v2-sparc64-u.out OLD_FILES+=usr/tests/usr.sbin/sa/v2-sparc64-usr.in -OLD_FILES+=usr/tests/usr.sbin/sa +OLD_DIRS+=usr/tests/usr.sbin/sa .endif .if ${MK_ACPI} == no ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r344136 - in stable/12: libexec/rc/rc.d tests/sys tests/sys/audit tools/build/mk
Author: asomers Date: Thu Feb 14 20:29:33 2019 New Revision: 344136 URL: https://svnweb.freebsd.org/changeset/base/344136 Log: MFC r342153, r342172-r342173 r342153: Conditionally install /etc/rc.d/audit* based on ${MK_AUDIT} /usr/sbin/audit(dist)?d are only installed if ${MK_AUDIT} == yes. Their supporting scripts should only be installed in those instances as well. Submitted by: ngie Reviewed by: emaste Pull Request: https://github.com/freebsd/freebsd/pull/242 r342172: audit(4) tests: require /etc/rc.d/auditd These tests should be skipped if /etc/rc.d/auditd is missing, which could be the case if world was built with WITHOUT_AUDIT set. Also, one test case requires /etc/rc.d/accounting. Submitted by: ngie Pull Request: https://github.com/freebsd/freebsd/pull/240 r342173: Conditionalize installtion audit(4) tests on MK_AUDIT MK_AUDIT already controls auditd(8), praudit(1), etc. It should also control the audit test suite. Submitted by: ngie Pull Request: https://github.com/freebsd/freebsd/pull/240 Modified: stable/12/libexec/rc/rc.d/Makefile stable/12/tests/sys/Makefile stable/12/tests/sys/audit/Makefile stable/12/tests/sys/audit/administrative.c stable/12/tools/build/mk/OptionalObsoleteFiles.inc Directory Properties: stable/12/ (props changed) Modified: stable/12/libexec/rc/rc.d/Makefile == --- stable/12/libexec/rc/rc.d/Makefile Thu Feb 14 20:26:59 2019 (r344135) +++ stable/12/libexec/rc/rc.d/Makefile Thu Feb 14 20:29:33 2019 (r344136) @@ -15,8 +15,6 @@ CONFS=DAEMON \ addswap \ adjkerntz \ archdep \ - auditd \ - auditdistd \ bgfsck \ ${_blacklistd} \ ${_bluetooth} \ @@ -162,6 +160,12 @@ APM+= apm APM+= apmd .endif APMPACKAGE=apm +.endif + +.if ${MK_AUDIT} != "no" +CONFGROUPS+= AUDIT +AUDIT+=auditd +AUDIT+=auditdistd .endif .if ${MK_AUTOFS} != "no" Modified: stable/12/tests/sys/Makefile == --- stable/12/tests/sys/MakefileThu Feb 14 20:26:59 2019 (r344135) +++ stable/12/tests/sys/MakefileThu Feb 14 20:29:33 2019 (r344136) @@ -6,7 +6,7 @@ TESTSDIR= ${TESTSBASE}/sys TESTS_SUBDIRS+=acl TESTS_SUBDIRS+=aio -TESTS_SUBDIRS+=audit +TESTS_SUBDIRS+=${_audit} TESTS_SUBDIRS+=auditpipe TESTS_SUBDIRS+=capsicum TESTS_SUBDIRS+=${_cddl} @@ -27,6 +27,10 @@ TESTS_SUBDIRS+= posixshm TESTS_SUBDIRS+=sys TESTS_SUBDIRS+=vfs TESTS_SUBDIRS+=vm + +.if ${MK_AUDIT} != "no" +_audit=audit +.endif .if ${MK_CDDL} != "no" _cddl= cddl Modified: stable/12/tests/sys/audit/Makefile == --- stable/12/tests/sys/audit/Makefile Thu Feb 14 20:26:59 2019 (r344135) +++ stable/12/tests/sys/audit/Makefile Thu Feb 14 20:29:33 2019 (r344136) @@ -49,6 +49,7 @@ SRCS.miscellaneous+= utils.c TEST_METADATA+= timeout="30" TEST_METADATA+= required_user="root" TEST_METADATA+= is_exclusive="true" +TEST_METADATA+=required_files="/etc/rc.d/auditd" WARNS?=6 Modified: stable/12/tests/sys/audit/administrative.c == --- stable/12/tests/sys/audit/administrative.c Thu Feb 14 20:26:59 2019 (r344135) +++ stable/12/tests/sys/audit/administrative.c Thu Feb 14 20:29:33 2019 (r344136) @@ -377,6 +377,8 @@ ATF_TC_HEAD(acct_success, tc) { atf_tc_set_md_var(tc, "descr", "Tests the audit of a successful " "acct(2) call"); + atf_tc_set_md_var(tc, "require.files", + "/etc/rc.d/accounting /etc/rc.d/auditd"); } ATF_TC_BODY(acct_success, tc) Modified: stable/12/tools/build/mk/OptionalObsoleteFiles.inc == --- stable/12/tools/build/mk/OptionalObsoleteFiles.inc Thu Feb 14 20:26:59 2019(r344135) +++ stable/12/tools/build/mk/OptionalObsoleteFiles.inc Thu Feb 14 20:29:33 2019(r344136) @@ -147,6 +147,8 @@ OLD_FILES+=usr/share/man/man3/unistruct.3.gz .endif .if ${MK_AUDIT} == no +OLD_FILES+=etc/rc.d/auditd +OLD_FILES+=etc/rc.d/auditdistd OLD_FILES+=usr/sbin/audit OLD_FILES+=usr/sbin/auditd OLD_FILES+=usr/sbin/auditdistd @@ -158,6 +160,22 @@ OLD_FILES+=usr/share/man/man5/auditdistd.conf.5.gz OLD_FILES+=usr/share/man/man8/audit.8.gz OLD_FILES+=usr/share/man/man8/auditd.8.gz OLD_FILES+=usr/share/man/man8/auditdistd.8.gz +OLD_FILES+=usr/tests/sys/audit/process-control
svn commit: r344135 - in stable/12/tests/sys/geom/class: . eli mirror
Author: asomers Date: Thu Feb 14 20:26:59 2019 New Revision: 344135 URL: https://svnweb.freebsd.org/changeset/base/344135 Log: MFC r341390, r341392, r341667 r341390: Remove some dead code from the geli tests This is detritus in the Makefile, leftover from 327662. r341392: Unbreak geli/gmirror testcases if their geom classes cannot be loaded The problem with the logic prior to this commit was twofold: 1. The wrong set of idioms (TAP-compatible) were being applied to the ATF testcases when run, resulting in confusing ATF failure results on setup. 2. The cleanup subroutines were broken when the geom classes could not be loaded as they exited with 0 unexpectedly. This commit changes the test code to source the class-specific configuration (conf.sh) once globally, instead of sourcing it per testcase and per cleanup subroutine, and to call the ATF-specific setup subroutine(s) inline in the testcases. The refactoring done is effectively a no-op for the TAP testcases, modulo any refactoring done to create common code between the ATF and TAP testcases. This unbreaks the geli testcases converted to ATF in r327662 and r327683, and the gmirror testcases added in r327780, respectively, when the geom class could not be loaded. tests/sys/geom/class/mirror/... While here, ignore errors when turning debug failpoint sysctl off, which could occur if the gmirror class was not loaded. Submitted by: ngie Pull Request: https://github.com/freebsd/freebsd/pull/241 r341667: geom tests: Fix cleanup of ATF tests since r341392 r341392 changed common test cleanup routines in a way that allowed them to be used by TAP tests as well as ATF tests. However, a late change made during code review resulted in cleanup being broken for ATF tests, which source geom_subr.sh separately during the body and cleanup phases of the test. The result was that md(4) devices wouldn't get cleaned up. X-MFC-With: 341392 Modified: stable/12/tests/sys/geom/class/eli/Makefile stable/12/tests/sys/geom/class/eli/attach_test.sh stable/12/tests/sys/geom/class/eli/conf.sh stable/12/tests/sys/geom/class/eli/configure_test.sh stable/12/tests/sys/geom/class/eli/delkey_test.sh stable/12/tests/sys/geom/class/eli/detach_test.sh stable/12/tests/sys/geom/class/eli/init_test.sh stable/12/tests/sys/geom/class/eli/integrity_test.sh stable/12/tests/sys/geom/class/eli/kill_test.sh stable/12/tests/sys/geom/class/eli/misc_test.sh stable/12/tests/sys/geom/class/eli/onetime_test.sh stable/12/tests/sys/geom/class/eli/resize_test.sh stable/12/tests/sys/geom/class/eli/setkey_test.sh stable/12/tests/sys/geom/class/geom_subr.sh stable/12/tests/sys/geom/class/mirror/sync_error.sh Directory Properties: stable/12/ (props changed) Modified: stable/12/tests/sys/geom/class/eli/Makefile == --- stable/12/tests/sys/geom/class/eli/Makefile Thu Feb 14 20:24:33 2019 (r344134) +++ stable/12/tests/sys/geom/class/eli/Makefile Thu Feb 14 20:26:59 2019 (r344135) @@ -21,10 +21,6 @@ ATF_TESTS_SH+= setkey_test ${PACKAGE}FILES+= conf.sh -.for t in ${TAP_TESTS_SH} -TEST_METADATA.$t+= required_user="root" -.endfor - CFLAGS.pbkdf2_test=-I${SRCTOP}/sys SRCS.pbkdf2_test= \ Modified: stable/12/tests/sys/geom/class/eli/attach_test.sh == --- stable/12/tests/sys/geom/class/eli/attach_test.sh Thu Feb 14 20:24:33 2019(r344134) +++ stable/12/tests/sys/geom/class/eli/attach_test.sh Thu Feb 14 20:26:59 2019(r344135) @@ -1,5 +1,7 @@ # $FreeBSD$ +. $(atf_get_srcdir)/conf.sh + atf_test_case attach_d cleanup attach_d_head() { @@ -8,7 +10,7 @@ attach_d_head() } attach_d_body() { - . $(atf_get_srcdir)/conf.sh + geli_test_setup sectors=100 md=$(attach_md -t malloc -s `expr $sectors + 1`) @@ -35,7 +37,6 @@ attach_d_body() } attach_d_cleanup() { - . $(atf_get_srcdir)/conf.sh geli_test_cleanup } @@ -47,7 +48,7 @@ attach_r_head() } attach_r_body() { - . $(atf_get_srcdir)/conf.sh + geli_test_setup sectors=100 md=$(attach_md -t malloc -s `expr $sectors + 1`) @@ -64,7 +65,6 @@ attach_r_body() } attach_r_cleanup() { - . $(atf_get_srcdir)/conf.sh geli_test_cleanup } @@ -76,7 +76,7 @@ nokey_head() } nokey_body() { - . $(atf_get_srcdir)/conf.sh + geli_test_setup sectors=100 md=$(attach_md -t malloc -s `expr $sectors + 1`) @@ -88,7 +88,6 @@ nokey_body() } nokey_cleanup() { - . $(atf_get_srcdir)/conf.sh geli_test_cleanup } Modified: stable/12/tests/sys/geom/class/eli/conf.sh == ---
svn commit: r344134 - stable/12/lib/libc/sys
Author: asomers Date: Thu Feb 14 20:24:33 2019 New Revision: 344134 URL: https://svnweb.freebsd.org/changeset/base/344134 Log: MFC r341598: stat(2): clarify which syscalls modify file timestamps The list of syscalls that modify st_atim, st_mtim, and st_ctim was quite out of date and probably not accurate to begin with. Update it, and make it clear that the list is open-ended. Reviewed by: kib Differential Revision:https://reviews.freebsd.org/D18410 Modified: stable/12/lib/libc/sys/stat.2 Directory Properties: stable/12/ (props changed) Modified: stable/12/lib/libc/sys/stat.2 == --- stable/12/lib/libc/sys/stat.2 Thu Feb 14 19:07:08 2019 (r344133) +++ stable/12/lib/libc/sys/stat.2 Thu Feb 14 20:24:33 2019 (r344134) @@ -28,7 +28,7 @@ .\" @(#)stat.2 8.4 (Berkeley) 5/1/95 .\" $FreeBSD$ .\" -.Dd December 1, 2017 +.Dd December 5, 2018 .Dt STAT 2 .Os .Sh NAME @@ -156,45 +156,53 @@ are: .Bl -tag -width ".Va st_birthtim" .It Va st_atim Time when file data was last accessed. -Changed by the -.Xr mknod 2 , -.Xr utimes 2 , +Changed implicitly by syscalls such as .Xr read 2 and -.Xr readv 2 -system calls. +.Xr readv 2 , +and explicitly by +.Xr utimes 2 . .It Va st_mtim Time when file data was last modified. -Changed by the +Changed implicitly by syscalls such as +.Xr truncate 2 , +.Xr write 2 , +and +.Xr writev 2 , +and explicitly by +.Xr utimes 2 . +Also, any syscall which modifies directory content changes the +.Va st_mtim +for the affected directory. +For instance, +.Xr creat 2 , .Xr mkdir 2 , -.Xr mkfifo 2 , -.Xr mknod 2 , -.Xr utimes 2 , -.Xr write 2 +.Xr rename 2 , +.Xr link 2 , and -.Xr writev 2 -system calls. +.Xr unlink 2 . .It Va st_ctim Time when file status was last changed (inode data modification). -Changed by the +Changed implicitly by any syscall that affects file metadata, including +.Va st_mtim , +such as .Xr chflags 2 , .Xr chmod 2 , .Xr chown 2 , +.Xr truncate 2 , +.Xr utimes 2 , +and +.Xr write 2 . +Also, any syscall which modifies directory content changes the +.Va st_ctim +for the affected directory. +For instance, .Xr creat 2 , -.Xr link 2 , .Xr mkdir 2 , -.Xr mkfifo 2 , -.Xr mknod 2 , .Xr rename 2 , -.Xr rmdir 2 , -.Xr symlink 2 , -.Xr truncate 2 , -.Xr unlink 2 , -.Xr utimes 2 , -.Xr write 2 +.Xr link 2 , and -.Xr writev 2 -system calls. +.Xr unlink 2 . .It Va st_birthtim Time when the inode was created. .El ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r344132 - head/sys/dev/ixl
On 2/14/19 10:22 AM, Rodney W. Grimes wrote: >> Author: erj >> Date: Thu Feb 14 18:02:37 2019 >> New Revision: 344132 >> URL: https://svnweb.freebsd.org/changeset/base/344132 >> >> Log: >> ixl: Fix panic caused by bug exposed by r344062 >> >> Don't use a struct if_irq for IFLIB_INTR_IOV type interrupts since that >> results >> in get_core_offset() being called on them, and get_core_offset() doesn't >> handle IFLIB_INTR_IOV type interrupts, which results in an assert() being >> triggered >> in iflib_irq_set_affinity(). >> >> PR:235730 >> Reported by: Jeffrey Pieper >> MFC after: 1 day > > Normally you would request an RE@ approval for a fast track to stable, > consider this message such an approval. That does not match our historical practice over the past 20 years. If we want to change that practice, that's a topic we can debate, but re@ has only required oversight on MFC's during slushes/freezes with the additional caveat of perhaps watching out for ABI breakage at any time (and requiring approvals for a known ABI breakage on a branch). -- John Baldwin ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r344129 - head
On Thu, Feb 14, 2019 at 12:00:22PM -0700, Warner Losh wrote: > On Thu, Feb 14, 2019 at 11:29 AM Rodney W. Grimes < > > > Differential Review: https://reviews.freebsd.org/D19193 > > > > You sited a differential, but not give any attribution > > to the external source :-( > > > > The differential review has that information. External contributors should be recognized by having their names appear in the commit logs. ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r344133 - head/sys/kern
Author: bde Date: Thu Feb 14 19:07:08 2019 New Revision: 344133 URL: https://svnweb.freebsd.org/changeset/base/344133 Log: Finish the fix for overflow in calcru1(). The previous fix was unnecessarily very slow up to 105 hours where the simple formula used previously worked, and unnecessarily slow by a factor of about 5/3 up to 388 days, and didn't work above 388 days. 388 days is not a long time, since it is a reasonable uptime, and for processes the times being calculated are aggregated over all threads, so with N CPUs running the same thread a runtime of 388 days is reachable after only 388 / N physical days. The PRs document overflow at 388 days, but don't try to fix it. Use the simple formula up to 76 hours. Then use a complicated general method that reduces to the simple formula up to a bit less than 105 hours, then reduces to the previous method without its extra work up to almost 388 days, then does more complicated reductions, usually many bits at a time so that this is not slow. This works up to half of maximum representable time (292271 years), with accumulated rounding errors of at most 32 usec. amd64 can do all this with no avoidable rounding errors in an inline asm with 2 instructions, but this is too special to use. __uint128_t can do the same with 100's of instructions on 64-bit arches. Long doubles with at least 64 bits of precision are the easiest method to use on i386 userland, but are hard to use in the kernel. PR: 76972 and duplicates Reviewed by: kib Modified: head/sys/kern/kern_resource.c Modified: head/sys/kern/kern_resource.c == --- head/sys/kern/kern_resource.c Thu Feb 14 18:02:37 2019 (r344132) +++ head/sys/kern/kern_resource.c Thu Feb 14 19:07:08 2019 (r344133) @@ -863,13 +863,88 @@ rufetchtd(struct thread *td, struct rusage *ru) calcru1(p, >td_rux, >ru_utime, >ru_stime); } +/* XXX: the MI version is too slow to use: */ +#ifndef __HAVE_INLINE_FLSLL +#defineflsll(x)(fls((x) >> 32) != 0 ? fls((x) >> 32) + 32 : fls(x)) +#endif + static uint64_t mul64_by_fraction(uint64_t a, uint64_t b, uint64_t c) { + uint64_t acc, bh, bl; + int i, s, sa, sb; + /* -* Compute floor(a * (b / c)) without overflowing, (b / c) <= 1.0. +* Calculate (a * b) / c accurately enough without overflowing. c +* must be nonzero, and its top bit must be 0. a or b must be +* <= c, and the implementation is tuned for b <= c. +* +* The comments about times are for use in calcru1() with units of +* microseconds for 'a' and stathz ticks at 128 Hz for b and c. +* +* Let n be the number of top zero bits in c. Each iteration +* either returns, or reduces b by right shifting it by at least n. +* The number of iterations is at most 1 + 64 / n, and the error is +* at most the number of iterations. +* +* It is very unusual to need even 2 iterations. Previous +* implementations overflowed essentially by returning early in the +* first iteration, with n = 38 giving overflow at 105+ hours and +* n = 32 giving overlow at at 388+ days despite a more careful +* calculation. 388 days is a reasonable uptime, and the calculation +* needs to work for the uptime times the number of CPUs since 'a' +* is per-process. */ - return ((a / c) * b + (a % c) * (b / c) + (a % c) * (b % c) / c); + if (a >= (uint64_t)1 << 63) + return (0); /* Unsupported arg -- can't happen. */ + acc = 0; + for (i = 0; i < 128; i++) { + sa = flsll(a); + sb = flsll(b); + if (sa + sb <= 64) + /* Up to 105 hours on first iteration. */ + return (acc + (a * b) / c); + if (a >= c) { + /* +* This reduction is based on a = q * c + r, with the +* remainder r < c. 'a' may be large to start, and +* moving bits from b into 'a' at the end of the loop +* sets the top bit of 'a', so the reduction makes +* significant progress. +*/ + acc += (a / c) * b; + a %= c; + sa = flsll(a); + if (sa + sb <= 64) + /* Up to 388 days on first iteration. */ + return (acc + (a * b) / c); + } + + /* +* This step writes a * b as a * ((bh << s) + bl) = +* a * (bh << s) + a * bl = (a << s) * bh + a * bl. The 2 +* additive terms are handled separately. Splitting in
Re: svn commit: r344129 - head
On Thu, Feb 14, 2019 at 11:29 AM Rodney W. Grimes < free...@pdx.rh.cn85.dnsmgr.net> wrote: > > Author: imp > > Date: Thu Feb 14 17:04:04 2019 > > New Revision: 344129 > > URL: https://svnweb.freebsd.org/changeset/base/344129 > > > > Log: > > Fix small typo. > > A small request, when we fix a typo could we > do: > fix typo: frambuffer -> framebuffer > No. Diff tells you that. > Thanks, > Rod > > > > Differential Review: https://reviews.freebsd.org/D19193 > > You sited a differential, but not give any attribution > to the external source :-( > The differential review has that information. Warner > > Modified: > > head/UPDATING > > > > Modified: head/UPDATING > > > == > > --- head/UPDATING Thu Feb 14 15:45:53 2019(r344128) > > +++ head/UPDATING Thu Feb 14 17:04:04 2019(r344129) > > @@ -237,7 +237,7 @@ NOTE TO PEOPLE WHO THINK THAT FreeBSD 13.x IS SLOW: > > 20180719: > > ARM64 now have efifb support, if you want to have serial console > > on your arm64 board when an screen is connected and the bootloader > > - setup a frambuffer for us to use, just add : > > + setup a framebuffer for us to use, just add : > > boot_serial=YES > > boot_multicons=YES > > in /boot/loader.conf > > -- > Rod Grimes > rgri...@freebsd.org > ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r344129 - head
> Author: imp > Date: Thu Feb 14 17:04:04 2019 > New Revision: 344129 > URL: https://svnweb.freebsd.org/changeset/base/344129 > > Log: > Fix small typo. A small request, when we fix a typo could we do: fix typo: frambuffer -> framebuffer Thanks, Rod > > Differential Review: https://reviews.freebsd.org/D19193 You sited a differential, but not give any attribution to the external source :-( > Modified: > head/UPDATING > > Modified: head/UPDATING > == > --- head/UPDATING Thu Feb 14 15:45:53 2019(r344128) > +++ head/UPDATING Thu Feb 14 17:04:04 2019(r344129) > @@ -237,7 +237,7 @@ NOTE TO PEOPLE WHO THINK THAT FreeBSD 13.x IS SLOW: > 20180719: > ARM64 now have efifb support, if you want to have serial console > on your arm64 board when an screen is connected and the bootloader > - setup a frambuffer for us to use, just add : > + setup a framebuffer for us to use, just add : > boot_serial=YES > boot_multicons=YES > in /boot/loader.conf -- Rod Grimes rgri...@freebsd.org ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r344132 - head/sys/dev/ixl
> Author: erj > Date: Thu Feb 14 18:02:37 2019 > New Revision: 344132 > URL: https://svnweb.freebsd.org/changeset/base/344132 > > Log: > ixl: Fix panic caused by bug exposed by r344062 > > Don't use a struct if_irq for IFLIB_INTR_IOV type interrupts since that > results > in get_core_offset() being called on them, and get_core_offset() doesn't > handle IFLIB_INTR_IOV type interrupts, which results in an assert() being > triggered > in iflib_irq_set_affinity(). > > PR: 235730 > Reported by:Jeffrey Pieper > MFC after: 1 day Normally you would request an RE@ approval for a fast track to stable, consider this message such an approval. > Sponsored by: Intel Corporation > > Modified: > head/sys/dev/ixl/if_ixl.c > head/sys/dev/ixl/ixl_pf.h > > Modified: head/sys/dev/ixl/if_ixl.c > == > --- head/sys/dev/ixl/if_ixl.c Thu Feb 14 18:01:06 2019(r344131) > +++ head/sys/dev/ixl/if_ixl.c Thu Feb 14 18:02:37 2019(r344132) > @@ -932,7 +932,7 @@ ixl_if_msix_intr_assign(if_ctx_t ctx, int msix) > return (err); > } > /* Create soft IRQ for handling VFLRs */ > - iflib_softirq_alloc_generic(ctx, >iov_irq, IFLIB_INTR_IOV, pf, 0, > "iov"); > + iflib_softirq_alloc_generic(ctx, NULL, IFLIB_INTR_IOV, pf, 0, "iov"); > > /* Now set up the stations */ > for (i = 0, vector = 1; i < vsi->shared->isc_nrxqsets; i++, vector++, > rx_que++) { > > Modified: head/sys/dev/ixl/ixl_pf.h > == > --- head/sys/dev/ixl/ixl_pf.h Thu Feb 14 18:01:06 2019(r344131) > +++ head/sys/dev/ixl/ixl_pf.h Thu Feb 14 18:02:37 2019(r344132) > @@ -138,7 +138,6 @@ struct ixl_pf { > struct ixl_vf *vfs; > int num_vfs; > uint16_tveb_seid; > - struct if_irq iov_irq; > }; > > /* -- Rod Grimes rgri...@freebsd.org ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r344132 - head/sys/dev/ixl
Author: erj Date: Thu Feb 14 18:02:37 2019 New Revision: 344132 URL: https://svnweb.freebsd.org/changeset/base/344132 Log: ixl: Fix panic caused by bug exposed by r344062 Don't use a struct if_irq for IFLIB_INTR_IOV type interrupts since that results in get_core_offset() being called on them, and get_core_offset() doesn't handle IFLIB_INTR_IOV type interrupts, which results in an assert() being triggered in iflib_irq_set_affinity(). PR: 235730 Reported by: Jeffrey Pieper MFC after:1 day Sponsored by: Intel Corporation Modified: head/sys/dev/ixl/if_ixl.c head/sys/dev/ixl/ixl_pf.h Modified: head/sys/dev/ixl/if_ixl.c == --- head/sys/dev/ixl/if_ixl.c Thu Feb 14 18:01:06 2019(r344131) +++ head/sys/dev/ixl/if_ixl.c Thu Feb 14 18:02:37 2019(r344132) @@ -932,7 +932,7 @@ ixl_if_msix_intr_assign(if_ctx_t ctx, int msix) return (err); } /* Create soft IRQ for handling VFLRs */ - iflib_softirq_alloc_generic(ctx, >iov_irq, IFLIB_INTR_IOV, pf, 0, "iov"); + iflib_softirq_alloc_generic(ctx, NULL, IFLIB_INTR_IOV, pf, 0, "iov"); /* Now set up the stations */ for (i = 0, vector = 1; i < vsi->shared->isc_nrxqsets; i++, vector++, rx_que++) { Modified: head/sys/dev/ixl/ixl_pf.h == --- head/sys/dev/ixl/ixl_pf.h Thu Feb 14 18:01:06 2019(r344131) +++ head/sys/dev/ixl/ixl_pf.h Thu Feb 14 18:02:37 2019(r344132) @@ -138,7 +138,6 @@ struct ixl_pf { struct ixl_vf *vfs; int num_vfs; uint16_tveb_seid; - struct if_irq iov_irq; }; /* ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r344131 - stable/12/sys/kern
Author: asomers Date: Thu Feb 14 18:01:06 2019 New Revision: 344131 URL: https://svnweb.freebsd.org/changeset/base/344131 Log: MFC r340988: vfs_aio.c: rename "physio" symbols to "bio". aio has two paths: an asynchronous "physio" path and a synchronous path. Confusingly, physio(9) isn't actually used by the "physio" path, and never has been. In fact, it may even be called by the synchronous path! Rename the "physio" path to the "bio" path to reflect what it actually does: directly compose BIOs and send them to character devices. Modified: stable/12/sys/kern/vfs_aio.c Directory Properties: stable/12/ (props changed) Modified: stable/12/sys/kern/vfs_aio.c == --- stable/12/sys/kern/vfs_aio.cThu Feb 14 17:40:38 2019 (r344130) +++ stable/12/sys/kern/vfs_aio.cThu Feb 14 18:01:06 2019 (r344131) @@ -212,7 +212,7 @@ typedef struct oaiocb { /* * If the routine that services an AIO request blocks while running in an * AIO kernel process it can starve other I/O requests. BIO requests - * queued via aio_qphysio() complete in GEOM and do not use AIO kernel + * queued via aio_qbio() complete asynchronously and do not use AIO kernel * processes at all. Socket I/O requests use a separate pool of * kprocs and also force non-blocking I/O. Other file I/O requests * use the generic fo_read/fo_write operations which can block. The @@ -268,7 +268,7 @@ struct kaioinfo { int kaio_flags; /* (a) per process kaio flags */ int kaio_active_count; /* (c) number of currently used AIOs */ int kaio_count; /* (a) size of AIO queue */ - int kaio_buffer_count; /* (a) number of physio buffers */ + int kaio_buffer_count; /* (a) number of bio buffers */ TAILQ_HEAD(,kaiocb) kaio_all; /* (a) all AIOs in a process */ TAILQ_HEAD(,kaiocb) kaio_done; /* (a) done queue for process */ TAILQ_HEAD(,aioliojob) kaio_liojoblist; /* (a) list of lio jobs */ @@ -318,11 +318,11 @@ static intaio_newproc(int *); intaio_aqueue(struct thread *td, struct aiocb *ujob, struct aioliojob *lio, int type, struct aiocb_ops *ops); static int aio_queue_file(struct file *fp, struct kaiocb *job); -static voidaio_physwakeup(struct bio *bp); +static voidaio_biowakeup(struct bio *bp); static voidaio_proc_rundown(void *arg, struct proc *p); static voidaio_proc_rundown_exec(void *arg, struct proc *p, struct image_params *imgp); -static int aio_qphysio(struct proc *p, struct kaiocb *job); +static int aio_qbio(struct proc *p, struct kaiocb *job); static voidaio_daemon(void *param); static voidaio_bio_done_notify(struct proc *userp, struct kaiocb *job); static boolaio_clear_cancel_function_locked(struct kaiocb *job); @@ -741,9 +741,9 @@ drop: /* * The AIO processing activity for LIO_READ/LIO_WRITE. This is the code that - * does the I/O request for the non-physio version of the operations. The - * normal vn operations are used, and this code should work in all instances - * for every type of file, including pipes, sockets, fifos, and regular files. + * does the I/O request for the non-bio version of the operations. The normal + * vn operations are used, and this code should work in all instances for every + * type of file, including pipes, sockets, fifos, and regular files. * * XXX I don't think it works well for socket, pipe, and fifo. */ @@ -1195,7 +1195,7 @@ aio_newproc(int *start) } /* - * Try the high-performance, low-overhead physio method for eligible + * Try the high-performance, low-overhead bio method for eligible * VCHR devices. This method doesn't use an aio helper thread, and * thus has very low overhead. * @@ -1204,7 +1204,7 @@ aio_newproc(int *start) * duration of this call. */ static int -aio_qphysio(struct proc *p, struct kaiocb *job) +aio_qbio(struct proc *p, struct kaiocb *job) { struct aiocb *cb; struct file *fp; @@ -1277,7 +1277,7 @@ aio_qphysio(struct proc *p, struct kaiocb *job) bp->bio_length = cb->aio_nbytes; bp->bio_bcount = cb->aio_nbytes; - bp->bio_done = aio_physwakeup; + bp->bio_done = aio_biowakeup; bp->bio_data = (void *)(uintptr_t)cb->aio_buf; bp->bio_offset = cb->aio_offset; bp->bio_cmd = cb->aio_lio_opcode == LIO_WRITE ? BIO_WRITE : BIO_READ; @@ -1442,7 +1442,7 @@ static struct aiocb_ops aiocb_ops_osigevent = { #endif /* - * Queue a new AIO request. Choosing either the threaded or direct physio VCHR + * Queue a new AIO request. Choosing either the threaded or direct bio VCHR * technique is done in this code. */ int @@ -1697,7 +1697,7 @@ aio_queue_file(struct file *fp, struct kaiocb *job) bool safe; ki = job->userproc->p_aioinfo; - error =
svn commit: r344130 - stable/12
Author: asomers Date: Thu Feb 14 17:40:38 2019 New Revision: 344130 URL: https://svnweb.freebsd.org/changeset/base/344130 Log: MFC r340525 again Apparently I didn't commit the SVN merge tracking info the first time I MFCed this commit. This commit is merge tracking info only. mount_fusefs.8: expand HISTORY section Note that fuse was available from ports long before joining the base system. Also, update the upstream URL. Modified: Directory Properties: stable/12/ (props changed) ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r344129 - head
Author: imp Date: Thu Feb 14 17:04:04 2019 New Revision: 344129 URL: https://svnweb.freebsd.org/changeset/base/344129 Log: Fix small typo. Differential Review: https://reviews.freebsd.org/D19193 Modified: head/UPDATING Modified: head/UPDATING == --- head/UPDATING Thu Feb 14 15:45:53 2019(r344128) +++ head/UPDATING Thu Feb 14 17:04:04 2019(r344129) @@ -237,7 +237,7 @@ NOTE TO PEOPLE WHO THINK THAT FreeBSD 13.x IS SLOW: 20180719: ARM64 now have efifb support, if you want to have serial console on your arm64 board when an screen is connected and the bootloader - setup a frambuffer for us to use, just add : + setup a framebuffer for us to use, just add : boot_serial=YES boot_multicons=YES in /boot/loader.conf ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r344114 - stable/12/sbin/sysctl
[ Charset UTF-8 unsupported, converting... ] > Author: ygy (doc committer) > Date: Thu Feb 14 09:12:19 2019 > New Revision: 344114 > URL: https://svnweb.freebsd.org/changeset/base/344114 > > Log: > MFC r343930: > > Remove -R option which was added to sysctl(8) man page per r244106, but it > is not implemented. > > Submitted by: Alfonso Siciliano > Reviewed by:0mp, imp > Differential Revision: https://reviews.freebsd.org/D19012 > > Modified: > stable/12/sbin/sysctl/sysctl.8 > Directory Properties: > stable/12/ (props changed) > > Modified: stable/12/sbin/sysctl/sysctl.8 > == > --- stable/12/sbin/sysctl/sysctl.8Thu Feb 14 00:52:03 2019 > (r344113) > +++ stable/12/sbin/sysctl/sysctl.8Thu Feb 14 09:12:19 2019 > (r344114) > @@ -28,7 +28,7 @@ > .\" From: @(#)sysctl.8 8.1 (Berkeley) 6/6/93 > .\" $FreeBSD$ > .\" > -.Dd September 24, 2018 > +.Dd February 8, 2019 > .Dt SYSCTL 8 > .Os > .Sh NAME > @@ -36,13 +36,13 @@ > .Nd get or set kernel state > .Sh SYNOPSIS > .Nm > -.Op Fl bdehiNnoRTtqx > +.Op Fl bdehiNnoTtqWx The commit message does not mention adding any flags, yet I see a new W flag here? This is a commit message error, as -W is in the man page and usage() of the program. > .Op Fl B Ar bufsize > .Op Fl f Ar filename > .Ar name Ns Op = Ns Ar value Ns Op , Ns Ar value > .Ar ... > .Nm > -.Op Fl bdehNnoRTtqx > +.Op Fl bdehNnoTtqWx And here as well. > .Op Fl B Ar bufsize > .Fl a > .Sh DESCRIPTION > > -- Rod Grimes rgri...@freebsd.org ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r343030 - in head/sys: cam conf dev/md dev/nvme fs/fuse fs/nfsclient fs/smbfs kern sys ufs/ffs vm
On Thu, Feb 14, 2019 at 06:56:42PM +1100, Bruce Evans wrote: > On Wed, 13 Feb 2019, Justin Hibbits wrote: > > > On Tue, 15 Jan 2019 01:02:17 + (UTC) > > Gleb Smirnoff wrote: > > > >> Author: glebius > >> Date: Tue Jan 15 01:02:16 2019 > >> New Revision: 343030 > >> URL: https://svnweb.freebsd.org/changeset/base/343030 > >> > >> Log: > >> Allocate pager bufs from UMA instead of 80-ish mutex protected > >> linked list. > > ... > > > > This seems to break 32-bit platforms, or at least 32-bit book-e > > powerpc, which has a limited KVA space (~500MB). It preallocates I've > > seen over 2500 pbufs, at 128kB each, eating up over 300MB KVA, > > leaving very little left for the rest of runtime. > > Hrmph. I complained other things in this commit this when it was > committed, but not this largest bug since preallocation was broken then > so I thought that it wasn't done, so that problems are smaller unless the > excessive limits are actually reached. > > Now i386 does it: > > XX ITEM SIZE LIMIT USED FREE REQ FAIL SLEEP > XX > XX swrbuf: 336,128, 0, 0, 0, 0, 0 > XX swwbuf: 336, 64, 0, 0, 0, 0, 0 > XX nfspbuf:336,128, 0, 0, 0, 0, 0 > XX mdpbuf: 336, 25, 0, 0, 0, 0, 0 > XX clpbuf: 336,128, 0, 5, 4, 0, 0 > XX vnpbuf: 336, 2048, 0, 0, 0, 0, 0 > XX pbuf: 336, 16, 0,2535, 0, 0, 0 > > but i386 now has 4GB of KVA, with almost 3GB to waste, so the bug is not > noticed there. > > The preallocation wasn't there in my last mail to the author about nearby > bugs, on 24 Jan 2019: > > YY vnpbuf: 568, 2048, 0, 0, 0, 0, 0 > YY clpbuf: 568,128, 0, 128,8750, 0, 1 > YY pbuf: 568, 16, 0, 4, 0, 0, 0 > > This output is on amd64 where the SIZE is larger and everything else was > the same as on i386. Now amd64 shows the large preallocation too. > > There seems to be another bug for the especially small LIMIT of 16 to > turn into a preallocation of 2535 and not cause immediate reduction to > the limit. > > I happen to have kernels from 24 and 25 Jan handy. The first one is > amd64 r343346M built on Jan 23, and it doesn't do the large > preallocation. The second one is i386 r343388:343418M built on Jan > 25, and it does the large preallocation. Both call uma_prealloc() to > ask for nswbuf_max = 0x9e9 buffers, but the old version only allocates > 4 buffers while later version allocate 0x9e9 buffers. > > The only relevant commit between the good and bad versions seems to be > r343453. This fixes uma_prealloc() to actually work. But it is a feature > for it to not work when its caller asks for too much. I guess you meant r343353. In any case, the pbuf keg is _NOFREE, so even without preallocation the large pbuf zone limits may become problematic if there are bursts of allocation requests. > 0x9e9 is the sum of the LIMITs of all pbuf pools. The main bug in > r343030 is that it expands nswbuf, which is supposed to give the > combined limit, from its normal value of 256 to 0x9e9. (r343030 > actually used nswbuf before it was properly initialized, so used its > maximum value of 256 even on small systems with nswbuf = 16. Only > this has been fixed.) > > On i386, nbuf is excessively limited so as to give a maxbufspace of > about 100MB so as to fit in 1GB of kva even with infinite RAM and > -current's actual 4GB of kva. nbuf is correctly limited to give a > much smaller maxbufspace when RAM is small (kva scaling for this is > not done so well). nswbuf is restricted if nbuf is restricted, but > not enough (except in my version). It is normally 256, so the pbuf > allocation used to be 32MB, and this is already a bit large compared > with 100MB for maxbufspace. Expanding pbufs by a factor of 0x9e9/0x100 > gives the silly combination of 100MB for maxbufspace and 317MB for > pbufs. > > If kva is only 512MB instead of 1GB, then maxbufspace should be only > 50MB and nswbuf should be smaller too. Similarly for PAE on i386 back > when it was configured with 1GB kva by default. Only about 512MB are > left after allocating space for page table metadata. I have fixes > that scale most of this better. Large subsystems starting with kmem > get a hard-coded fraction of the usable kva. E.g., kmem gets about > 60% of usable kva instead of about 40% of nominal kva. Most other > large subsystems including the buffer cache get about 1/8 of the > remaining 40% of usable kva. Scaling for other subsystems is mostly > worse than for kmem. pbufs are part of the buffer cache allocation. > The expansion factor of 0x9e9/0x100 breaks this. > > I don't understand how pbuf_preallocate() allocates
svn commit: r344128 - head/sys/vm
Author: kib Date: Thu Feb 14 15:45:53 2019 New Revision: 344128 URL: https://svnweb.freebsd.org/changeset/base/344128 Log: Make anon clustering more compatible. Make the clustering enabling knob more fine-grained by providing a setting where the allocation with hint is not clustered. This is aimed to be somewhat more compatible with e.g. go 1.4 which expects that hinted mmap without MAP_FIXED does not change the allocation address. Now the vm.cluster_anon can be set to 1 to only cluster when no hints, and to 2 to always cluster. Default value is 1. Requested by: peter Reviewed by: emaste, markj Sponsored by: The FreeBSD Foundation MFC after:1 month Differential revision:https://reviews.freebsd.org/D19194 Modified: head/sys/vm/vm_map.c Modified: head/sys/vm/vm_map.c == --- head/sys/vm/vm_map.cThu Feb 14 15:42:29 2019(r344127) +++ head/sys/vm/vm_map.cThu Feb 14 15:45:53 2019(r344128) @@ -1487,8 +1487,23 @@ static const int aslr_pages_rnd_32[2] = {0x100, 0x4}; static int cluster_anon = 1; SYSCTL_INT(_vm, OID_AUTO, cluster_anon, CTLFLAG_RW, _anon, 0, -"Cluster anonymous mappings"); +"Cluster anonymous mappings: 0 = no, 1 = yes if no hint, 2 = always"); +static bool +clustering_anon_allowed(vm_offset_t addr) +{ + + switch (cluster_anon) { + case 0: + return (false); + case 1: + return (addr == 0); + case 2: + default: + return (true); + } +} + static long aslr_restarts; SYSCTL_LONG(_vm, OID_AUTO, aslr_restarts, CTLFLAG_RD, _restarts, 0, @@ -1593,7 +1608,7 @@ vm_map_find(vm_map_t map, vm_object_t object, vm_ooffs } else alignment = 0; en_aslr = (map->flags & MAP_ASLR) != 0; - update_anon = cluster = cluster_anon != 0 && + update_anon = cluster = clustering_anon_allowed(*addr) && (map->flags & MAP_IS_SUB_MAP) == 0 && max_addr == 0 && find_space != VMFS_NO_SPACE && object == NULL && (cow & (MAP_INHERIT_SHARE | MAP_STACK_GROWS_UP | ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r344127 - stable/11/lib/libfigpar
Author: se Date: Thu Feb 14 15:42:29 2019 New Revision: 344127 URL: https://svnweb.freebsd.org/changeset/base/344127 Log: MFC r343480,343482: Silence Clang Scan warning about unsafe use of strcpy. Replace strcpy() by memcpy to the previously allocated range of known size. Modified: stable/11/lib/libfigpar/string_m.c Directory Properties: stable/11/ (props changed) Modified: stable/11/lib/libfigpar/string_m.c == --- stable/11/lib/libfigpar/string_m.c Thu Feb 14 15:41:05 2019 (r344126) +++ stable/11/lib/libfigpar/string_m.c Thu Feb 14 15:42:29 2019 (r344127) @@ -120,9 +120,9 @@ replaceall(char *source, const char *find, const char /* If replace is longer than find, we'll need to create a temp copy */ if (rlen > flen) { temp = malloc(slen + 1); - if (errno != 0) /* could not allocate memory */ + if (temp == NULL) /* could not allocate memory */ return (-1); - strcpy(temp, source); + memcpy(temp, source, slen + 1); } else temp = source; ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r344126 - stable/11/libexec/getty
Author: se Date: Thu Feb 14 15:41:05 2019 New Revision: 344126 URL: https://svnweb.freebsd.org/changeset/base/344126 Log: MFC r343479: Fix potential buffer overflow and undefined behavior. The buffer allocated in read_chat() could be 1 element too short, if the chatstr parameter passed in is 1 or 3 charachters long (e.g. "a" or "a b"). The allocation of the pointer array does not account for the terminating NULL pointer in that case. Overlapping source and destination strings are undefined in strcpy(). Instead of moving a string to the left by one character just increment the char pointer before it is assigned to the results array. Modified: stable/11/libexec/getty/chat.c Directory Properties: stable/11/ (props changed) Modified: stable/11/libexec/getty/chat.c == --- stable/11/libexec/getty/chat.c Thu Feb 14 15:39:17 2019 (r344125) +++ stable/11/libexec/getty/chat.c Thu Feb 14 15:41:05 2019 (r344126) @@ -141,7 +141,7 @@ read_chat(char **chatstr) int l; if ((l=strlen(str)) > 0 && (tmp=malloc(l + 1)) != NULL && - (res=malloc((l / 2 + 1) * sizeof(char *))) != NULL) { + (res=malloc(((l + 1) / 2 + 1) * sizeof(char *))) != NULL) { static char ws[] = " \t"; char * p; @@ -216,7 +216,7 @@ read_chat(char **chatstr) q = strrchr(p+1, *p); if (q != NULL && *q == *p && q[1] == '\0') { *q = '\0'; - strcpy(p, p+1); + p++; } } ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r344125 - stable/11/usr.bin/whereis
Author: se Date: Thu Feb 14 15:39:17 2019 New Revision: 344125 URL: https://svnweb.freebsd.org/changeset/base/344125 Log: MFC r343408: Silence Clang Scan warnings regarding unsafe use of strcp(). While these warnings are false positives, the use of strdup() instead of malloc() and strcpy() simplifies and clarifies the code. A bogus assignment to a variable (whose previous value may be required in a later block) has also been removed. Modified: stable/11/usr.bin/whereis/whereis.c Directory Properties: stable/11/ (props changed) Modified: stable/11/usr.bin/whereis/whereis.c == --- stable/11/usr.bin/whereis/whereis.c Thu Feb 14 15:33:04 2019 (r344124) +++ stable/11/usr.bin/whereis/whereis.c Thu Feb 14 15:39:17 2019 (r344125) @@ -285,9 +285,9 @@ defaults(void) bindirs[nele] = NULL; if ((cp = getenv("PATH")) != NULL) { /* don't destroy the original environment... */ - if ((b = malloc(strlen(cp) + 1)) == NULL) + b = strdup(cp); + if (b == NULL) abort(); - strcpy(b, cp); decolonify(b, , ); } } @@ -301,18 +301,18 @@ defaults(void) err(EX_OSERR, "error processing manpath results"); if ((b = strchr(buf, '\n')) != NULL) *b = '\0'; - if ((b = malloc(strlen(buf) + 1)) == NULL) + b = strdup(buf); + if (b == NULL) abort(); - strcpy(b, buf); nele = 0; decolonify(b, , ); } /* -s defaults to precompiled list, plus subdirs of /usr/ports */ if (!sourcedirs) { - if ((b = malloc(strlen(sourcepath) + 1)) == NULL) + b = strdup(sourcepath); + if (b == NULL) abort(); - strcpy(b, sourcepath); nele = 0; decolonify(b, , ); @@ -523,11 +523,9 @@ main(int argc, char **argv) * man -w found plain source * page, use it. */ - s = strlen(buf); - cp2 = malloc(s + 1); + cp2 = strdup(buf); if (cp2 == NULL) abort(); - strcpy(cp2, buf); } if (man == NULL) { ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r344124 - stable/11/usr.sbin/kbdcontrol
Author: se Date: Thu Feb 14 15:33:04 2019 New Revision: 344124 URL: https://svnweb.freebsd.org/changeset/base/344124 Log: MFC r343339: Silence Clang Scan warning about use of unitialized variable. The logic is changed to depend on actual "beep" parameters instead of on a flag that may be set for invalid parameters. An embedded literal escape character has been replaced by "\e", since it could confuse terminals when displaying the affected line. Modified: stable/11/usr.sbin/kbdcontrol/kbdcontrol.c Directory Properties: stable/11/ (props changed) Modified: stable/11/usr.sbin/kbdcontrol/kbdcontrol.c == --- stable/11/usr.sbin/kbdcontrol/kbdcontrol.c Thu Feb 14 15:15:32 2019 (r344123) +++ stable/11/usr.sbin/kbdcontrol/kbdcontrol.c Thu Feb 14 15:33:04 2019 (r344124) @@ -961,6 +961,8 @@ set_bell_values(char *opt) int bell, duration, pitch; bell = 0; + duration = 0; + pitch = 0; if (!strncmp(opt, "quiet.", 6)) { bell = CONS_QUIET_BELL; opt += 6; @@ -991,8 +993,8 @@ badopt: } ioctl(0, CONS_BELLTYPE, ); - if (!(bell & CONS_VISUAL_BELL)) - fprintf(stderr, "[=%d;%dB", pitch, duration); + if (duration > 0 && pitch > 0) + fprintf(stderr, "\e[=%d;%dB", pitch, duration); } static void ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r344123 - head/sys/powerpc/powerpc
Author: luporl Date: Thu Feb 14 15:15:32 2019 New Revision: 344123 URL: https://svnweb.freebsd.org/changeset/base/344123 Log: [PPC64] Fix mismatch between thread flags and MSR When sigreturn() restored a thread's context, SRR1 was being restored to its previous value, but pcb_flags was not being touched. This could cause a mismatch between the thread's MSR and its pcb_flags. For instance, when the thread used the FPU for the first time inside the signal handler, sigreturn() would clear SRR1, but not pcb_flags. Then, the thread would return with the FPU bit cleared in MSR and, the next time it tried to use the FPU, it would fail on a KASSERT that checked if the FPU was disabled. This change clears the FPU bit in both pcb_flags and frame->srr1, as the code that restores the context expects to use the FPU trap to re-enable it. PR: 234539 Reported by: sbruno Reviewed by: jhibbits, sbruno Differential Revision:https://reviews.freebsd.org/D19166 Modified: head/sys/powerpc/powerpc/exec_machdep.c Modified: head/sys/powerpc/powerpc/exec_machdep.c == --- head/sys/powerpc/powerpc/exec_machdep.c Thu Feb 14 14:50:47 2019 (r344122) +++ head/sys/powerpc/powerpc/exec_machdep.c Thu Feb 14 15:15:32 2019 (r344123) @@ -474,6 +474,10 @@ set_mcontext(struct thread *td, mcontext_t *mcp) else tf->fixreg[2] = tls; + /* Disable FPU */ + tf->srr1 &= ~PSL_FP; + pcb->pcb_flags &= ~PCB_FPU; + if (mcp->mc_flags & _MC_FP_VALID) { /* enable_fpu() will happen lazily on a fault */ pcb->pcb_flags |= PCB_FPREGS; ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r344051 - in stable/12/sbin: newfs tunefs
On Tue, Feb 12, 2019 at 08:31:52AM -0800, Rodney W. Grimes wrote: > If all of our process was better documented and laid down > we would have less of this type of discussion on a much less > frequent basis. And much more discussion about how the process is over-specified, onerous, doesn't take care of every edge case, and so forth. I speak from personal knowledge of trying to deal with this during my several years on portmgr@. IMVVHO rules-lawyering is toxic to a volunteer project. It was enough to cause me to walk away from what I had taken on. mcl ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r344122 - stable/12/sys/vm
Author: kib Date: Thu Feb 14 14:50:47 2019 New Revision: 344122 URL: https://svnweb.freebsd.org/changeset/base/344122 Log: MFC r343850: contigmalloc: handle M_EXEC. Modified: stable/12/sys/vm/vm_kern.c Directory Properties: stable/12/ (props changed) Modified: stable/12/sys/vm/vm_kern.c == --- stable/12/sys/vm/vm_kern.c Thu Feb 14 14:44:53 2019(r344121) +++ stable/12/sys/vm/vm_kern.c Thu Feb 14 14:50:47 2019(r344122) @@ -184,6 +184,7 @@ kmem_alloc_attr_domain(int domain, vm_size_t size, int vm_offset_t addr, i, offset; vm_page_t m; int pflags, tries; + vm_prot_t prot; size = round_page(size); vmem = vm_dom[domain].vmd_kernel_arena; @@ -193,6 +194,7 @@ kmem_alloc_attr_domain(int domain, vm_size_t size, int pflags = malloc2vm_flags(flags) | VM_ALLOC_NOBUSY | VM_ALLOC_WIRED; pflags &= ~(VM_ALLOC_NOWAIT | VM_ALLOC_WAITOK | VM_ALLOC_WAITFAIL); pflags |= VM_ALLOC_NOWAIT; + prot = (flags & M_EXEC) != 0 ? VM_PROT_ALL : VM_PROT_RW; VM_OBJECT_WLOCK(object); for (i = 0; i < size; i += PAGE_SIZE) { tries = 0; @@ -220,8 +222,8 @@ retry: if ((flags & M_ZERO) && (m->flags & PG_ZERO) == 0) pmap_zero_page(m); m->valid = VM_PAGE_BITS_ALL; - pmap_enter(kernel_pmap, addr + i, m, VM_PROT_RW, - VM_PROT_RW | PMAP_ENTER_WIRED, 0); + pmap_enter(kernel_pmap, addr + i, m, prot, + prot | PMAP_ENTER_WIRED, 0); } VM_OBJECT_WUNLOCK(object); return (addr); ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r344121 - in head/sys: arm/arm arm64/arm64 mips/mips powerpc/powerpc riscv/riscv sparc64/sparc64
Author: kib Date: Thu Feb 14 14:44:53 2019 New Revision: 344121 URL: https://svnweb.freebsd.org/changeset/base/344121 Log: Enable enabling ASLR on non-x86 architectures. Discussed with: emaste Sponsored by: The FreeBSD Foundation Modified: head/sys/arm/arm/elf_machdep.c head/sys/arm64/arm64/elf_machdep.c head/sys/mips/mips/elf_machdep.c head/sys/powerpc/powerpc/elf32_machdep.c head/sys/powerpc/powerpc/elf64_machdep.c head/sys/riscv/riscv/elf_machdep.c head/sys/sparc64/sparc64/elf_machdep.c Modified: head/sys/arm/arm/elf_machdep.c == --- head/sys/arm/arm/elf_machdep.c Thu Feb 14 14:02:33 2019 (r344120) +++ head/sys/arm/arm/elf_machdep.c Thu Feb 14 14:44:53 2019 (r344121) @@ -84,7 +84,7 @@ struct sysentvec elf32_freebsd_sysvec = { #if __ARM_ARCH >= 6 SV_ASLR | SV_SHP | SV_TIMEKEEP | #endif - SV_ABI_FREEBSD | SV_ILP32, + SV_ABI_FREEBSD | SV_ILP32 | SV_ASLR, .sv_set_syscall_retval = cpu_set_syscall_retval, .sv_fetch_syscall_args = cpu_fetch_syscall_args, .sv_syscallnames = syscallnames, Modified: head/sys/arm64/arm64/elf_machdep.c == --- head/sys/arm64/arm64/elf_machdep.c Thu Feb 14 14:02:33 2019 (r344120) +++ head/sys/arm64/arm64/elf_machdep.c Thu Feb 14 14:44:53 2019 (r344121) @@ -79,7 +79,8 @@ static struct sysentvec elf64_freebsd_sysvec = { .sv_setregs = exec_setregs, .sv_fixlimit= NULL, .sv_maxssiz = NULL, - .sv_flags = SV_SHP | SV_TIMEKEEP | SV_ABI_FREEBSD | SV_LP64, + .sv_flags = SV_SHP | SV_TIMEKEEP | SV_ABI_FREEBSD | SV_LP64 | + SV_ASLR, .sv_set_syscall_retval = cpu_set_syscall_retval, .sv_fetch_syscall_args = cpu_fetch_syscall_args, .sv_syscallnames = syscallnames, Modified: head/sys/mips/mips/elf_machdep.c == --- head/sys/mips/mips/elf_machdep.cThu Feb 14 14:02:33 2019 (r344120) +++ head/sys/mips/mips/elf_machdep.cThu Feb 14 14:44:53 2019 (r344121) @@ -76,7 +76,7 @@ struct sysentvec elf64_freebsd_sysvec = { .sv_setregs = exec_setregs, .sv_fixlimit= NULL, .sv_maxssiz = NULL, - .sv_flags = SV_ABI_FREEBSD | SV_LP64, + .sv_flags = SV_ABI_FREEBSD | SV_LP64 | SV_ASLR, .sv_set_syscall_retval = cpu_set_syscall_retval, .sv_fetch_syscall_args = cpu_fetch_syscall_args, .sv_syscallnames = syscallnames, @@ -131,7 +131,7 @@ struct sysentvec elf32_freebsd_sysvec = { .sv_setregs = exec_setregs, .sv_fixlimit= NULL, .sv_maxssiz = NULL, - .sv_flags = SV_ABI_FREEBSD | SV_ILP32, + .sv_flags = SV_ABI_FREEBSD | SV_ILP32 | SV_ASLR, .sv_set_syscall_retval = cpu_set_syscall_retval, .sv_fetch_syscall_args = cpu_fetch_syscall_args, .sv_syscallnames = syscallnames, Modified: head/sys/powerpc/powerpc/elf32_machdep.c == --- head/sys/powerpc/powerpc/elf32_machdep.cThu Feb 14 14:02:33 2019 (r344120) +++ head/sys/powerpc/powerpc/elf32_machdep.cThu Feb 14 14:44:53 2019 (r344121) @@ -115,7 +115,7 @@ struct sysentvec elf32_freebsd_sysvec = { .sv_fixlimit= NULL, #endif .sv_maxssiz = NULL, - .sv_flags = SV_ABI_FREEBSD | SV_ILP32 | SV_SHP, + .sv_flags = SV_ABI_FREEBSD | SV_ILP32 | SV_SHP | SV_ASLR, .sv_set_syscall_retval = cpu_set_syscall_retval, .sv_fetch_syscall_args = cpu_fetch_syscall_args, .sv_shared_page_base = FREEBSD32_SHAREDPAGE, Modified: head/sys/powerpc/powerpc/elf64_machdep.c == --- head/sys/powerpc/powerpc/elf64_machdep.cThu Feb 14 14:02:33 2019 (r344120) +++ head/sys/powerpc/powerpc/elf64_machdep.cThu Feb 14 14:44:53 2019 (r344121) @@ -79,7 +79,7 @@ struct sysentvec elf64_freebsd_sysvec_v1 = { .sv_setregs = exec_setregs_funcdesc, .sv_fixlimit= NULL, .sv_maxssiz = NULL, - .sv_flags = SV_ABI_FREEBSD | SV_LP64 | SV_SHP, + .sv_flags = SV_ABI_FREEBSD | SV_LP64 | SV_SHP | SV_ASLR, .sv_set_syscall_retval = cpu_set_syscall_retval, .sv_fetch_syscall_args = cpu_fetch_syscall_args, .sv_syscallnames = syscallnames, Modified: head/sys/riscv/riscv/elf_machdep.c == --- head/sys/riscv/riscv/elf_machdep.c Thu Feb 14 14:02:33 2019 (r344120) +++ head/sys/riscv/riscv/elf_machdep.c Thu Feb 14 14:44:53 2019 (r344121) @@
svn commit: r344120 - in head/lib/libc: . amd64/gen i386/gen x86/gen
Author: kib Date: Thu Feb 14 14:02:33 2019 New Revision: 344120 URL: https://svnweb.freebsd.org/changeset/base/344120 Log: Unify i386 and amd64 getcontextx.c, and use ifuncs while there. In particular, use ifuncs for __getcontextx_size(), also calculate the size of the extended save area in resolver. Same for __fillcontextx2(). Sponsored by: The FreeBSD Foundation MFC after:1 week Added: head/lib/libc/x86/gen/ head/lib/libc/x86/gen/Makefile.inc (contents, props changed) head/lib/libc/x86/gen/getcontextx.c - copied, changed from r344117, head/lib/libc/amd64/gen/getcontextx.c Deleted: head/lib/libc/amd64/gen/getcontextx.c head/lib/libc/i386/gen/getcontextx.c Modified: head/lib/libc/Makefile head/lib/libc/amd64/gen/Makefile.inc head/lib/libc/i386/gen/Makefile.inc Modified: head/lib/libc/Makefile == --- head/lib/libc/Makefile Thu Feb 14 13:59:00 2019(r344119) +++ head/lib/libc/Makefile Thu Feb 14 14:02:33 2019(r344120) @@ -122,6 +122,7 @@ NOASM= .endif .if ${LIBC_ARCH} == "i386" || ${LIBC_ARCH} == "amd64" .include "${LIBC_SRCTOP}/x86/sys/Makefile.inc" +.include "${LIBC_SRCTOP}/x86/gen/Makefile.inc" .endif .if ${MK_NIS} != "no" CFLAGS+= -DYP Modified: head/lib/libc/amd64/gen/Makefile.inc == --- head/lib/libc/amd64/gen/Makefile.incThu Feb 14 13:59:00 2019 (r344119) +++ head/lib/libc/amd64/gen/Makefile.incThu Feb 14 14:02:33 2019 (r344120) @@ -2,7 +2,7 @@ # $FreeBSD$ SRCS+= _setjmp.S _set_tp.c rfork_thread.S setjmp.S sigsetjmp.S \ - fabs.S getcontextx.c \ + fabs.S \ infinity.c ldexp.c makecontext.c signalcontext.c \ flt_rounds.c fpgetmask.c fpsetmask.c fpgetprec.c fpsetprec.c \ fpgetround.c fpsetround.c fpgetsticky.c Modified: head/lib/libc/i386/gen/Makefile.inc == --- head/lib/libc/i386/gen/Makefile.inc Thu Feb 14 13:59:00 2019 (r344119) +++ head/lib/libc/i386/gen/Makefile.inc Thu Feb 14 14:02:33 2019 (r344120) @@ -2,5 +2,5 @@ # $FreeBSD$ SRCS+= _ctx_start.S _setjmp.S _set_tp.c fabs.S \ - flt_rounds.c getcontextx.c infinity.c ldexp.c makecontext.c \ + flt_rounds.c infinity.c ldexp.c makecontext.c \ rfork_thread.S setjmp.S signalcontext.c sigsetjmp.S Added: head/lib/libc/x86/gen/Makefile.inc == --- /dev/null 00:00:00 1970 (empty, because file is newly added) +++ head/lib/libc/x86/gen/Makefile.inc Thu Feb 14 14:02:33 2019 (r344120) @@ -0,0 +1,6 @@ +# $FreeBSD$ + +.PATH: ${LIBC_SRCTOP}/x86/gen + +SRCS+= \ + getcontextx.c Copied and modified: head/lib/libc/x86/gen/getcontextx.c (from r344117, head/lib/libc/amd64/gen/getcontextx.c) == --- head/lib/libc/amd64/gen/getcontextx.c Thu Feb 14 09:50:59 2019 (r344117, copy source) +++ head/lib/libc/x86/gen/getcontextx.c Thu Feb 14 14:02:33 2019 (r344120) @@ -35,49 +35,76 @@ __FBSDID("$FreeBSD$"); #include #include #include -#include #include #include +#include +#include -static int xstate_sz = -1; +#if defined __i386__ +#defineX86_GET_XFPUSTATE I386_GET_XFPUSTATE +typedef struct savexmm savex86_t ; +typedef struct i386_get_xfpustate x86_get_xfpustate_t; +#elif defined __amd64__ +#defineX86_GET_XFPUSTATE AMD64_GET_XFPUSTATE +typedef struct savefpu savex86_t; +typedef struct amd64_get_xfpustate x86_get_xfpustate_t; +#else +#error "Wrong arch" +#endif -int -__getcontextx_size(void) +static int xstate_sz = 0; + +static int +__getcontextx_size_xfpu(void) { + + return (sizeof(ucontext_t) + xstate_sz); +} + +DEFINE_UIFUNC(, int, __getcontextx_size, (void), static) +{ u_int p[4]; - if (xstate_sz == -1) { - do_cpuid(1, p); - if ((p[2] & CPUID2_OSXSAVE) != 0) { - cpuid_count(0xd, 0x0, p); - xstate_sz = p[1] - sizeof(struct savefpu); - } else - xstate_sz = 0; + if ((cpu_feature2 & CPUID2_OSXSAVE) != 0) { + cpuid_count(0xd, 0x0, p); + xstate_sz = p[1] - sizeof(savex86_t); } + return (__getcontextx_size_xfpu); +} - return (sizeof(ucontext_t) + xstate_sz); +static int +__fillcontextx2_xfpu(char *ctx) +{ + x86_get_xfpustate_t xfpu; + ucontext_t *ucp; + + ucp = (ucontext_t *)ctx; + xfpu.addr = (char *)(ucp + 1); + xfpu.len = xstate_sz; + if (sysarch(X86_GET_XFPUSTATE, ) == -1) + return (-1); + ucp->uc_mcontext.mc_xfpustate = (__register_t)xfpu.addr; + ucp->uc_mcontext.mc_xfpustate_len =
svn commit: r344119 - head/lib/libc/x86/sys
Author: kib Date: Thu Feb 14 13:59:00 2019 New Revision: 344119 URL: https://svnweb.freebsd.org/changeset/base/344119 Log: x86 __vdso_gettc(): use machine/cpufunc.h function for CPUID. Based on the discussion with: jkim Sponsored by: The FreeBSD Foundation MFC after:1 week Modified: head/lib/libc/x86/sys/__vdso_gettc.c Modified: head/lib/libc/x86/sys/__vdso_gettc.c == --- head/lib/libc/x86/sys/__vdso_gettc.cThu Feb 14 13:53:11 2019 (r344118) +++ head/lib/libc/x86/sys/__vdso_gettc.cThu Feb 14 13:59:00 2019 (r344119) @@ -54,31 +54,6 @@ __FBSDID("$FreeBSD$"); #include "libc_private.h" static void -cpuidp(u_int leaf, u_int p[4]) -{ - - __asm __volatile( -#if defined(__i386__) - " pushl %%ebx\n" -#endif - " cpuid\n" -#if defined(__i386__) - " movl%%ebx,%1\n" - " popl%%ebx" -#endif - : "=a" (p[0]), -#if defined(__i386__) - "=r" (p[1]), -#elif defined(__amd64__) - "=b" (p[1]), -#else -#error "Arch" -#endif - "=c" (p[2]), "=d" (p[3]) - : "0" (leaf)); -} - -static void rdtsc_mb_lfence(void) { @@ -100,12 +75,12 @@ rdtsc_mb_none(void) DEFINE_UIFUNC(static, void, rdtsc_mb, (void), static) { u_int p[4]; - /* Not a typo, string matches our cpuidp() registers use. */ + /* Not a typo, string matches our do_cpuid() registers use. */ static const char intel_id[] = "GenuntelineI"; if ((cpu_feature & CPUID_SSE2) == 0) return (rdtsc_mb_none); - cpuidp(0, p); + do_cpuid(0, p); return (memcmp(p + 1, intel_id, sizeof(intel_id) - 1) == 0 ? rdtsc_mb_lfence : rdtsc_mb_mfence); } ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r344118 - head/sys/i386/include
Author: kib Date: Thu Feb 14 13:53:11 2019 New Revision: 344118 URL: https://svnweb.freebsd.org/changeset/base/344118 Log: Provide userspace versions of do_cpuid() and cpuid_count() on i386. Some older compilers, when generating PIC code, cannot handle inline asm that clobbers %ebx (because %ebx is used as the GOT offset register). Userspace versions avoid clobbering %ebx by saving it to stack before executing the CPUID instruction. Sponsored by: The FreeBSD Foundation MFC after:1 week Modified: head/sys/i386/include/cpufunc.h Modified: head/sys/i386/include/cpufunc.h == --- head/sys/i386/include/cpufunc.h Thu Feb 14 09:50:59 2019 (r344117) +++ head/sys/i386/include/cpufunc.h Thu Feb 14 13:53:11 2019 (r344118) @@ -108,21 +108,47 @@ disable_intr(void) __asm __volatile("cli" : : : "memory"); } +#ifdef _KERNEL static __inline void do_cpuid(u_int ax, u_int *p) { __asm __volatile("cpuid" -: "=a" (p[0]), "=b" (p[1]), "=c" (p[2]), "=d" (p[3]) -: "0" (ax)); + : "=a" (p[0]), "=b" (p[1]), "=c" (p[2]), "=d" (p[3]) + : "0" (ax)); } static __inline void cpuid_count(u_int ax, u_int cx, u_int *p) { __asm __volatile("cpuid" -: "=a" (p[0]), "=b" (p[1]), "=c" (p[2]), "=d" (p[3]) -: "0" (ax), "c" (cx)); + : "=a" (p[0]), "=b" (p[1]), "=c" (p[2]), "=d" (p[3]) + : "0" (ax), "c" (cx)); } +#else +static __inline void +do_cpuid(u_int ax, u_int *p) +{ + __asm __volatile( + "pushl\t%%ebx\n\t" + "cpuid\n\t" + "movl\t%%ebx,%1\n\t" + "popl\t%%ebx" + : "=a" (p[0]), "=DS" (p[1]), "=c" (p[2]), "=d" (p[3]) + : "0" (ax)); +} + +static __inline void +cpuid_count(u_int ax, u_int cx, u_int *p) +{ + __asm __volatile( + "pushl\t%%ebx\n\t" + "cpuid\n\t" + "movl\t%%ebx,%1\n\t" + "popl\t%%ebx" + : "=a" (p[0]), "=DS" (p[1]), "=c" (p[2]), "=d" (p[3]) + : "0" (ax), "c" (cx)); +} +#endif static __inline void enable_intr(void) ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r344117 - stable/11/tools/build/mk
Author: avos Date: Thu Feb 14 09:50:59 2019 New Revision: 344117 URL: https://svnweb.freebsd.org/changeset/base/344117 Log: MFC r343868: Add ypldap(8)-related files and /var/yp directory to the OptionalObsoleteFiles.inc. Modified: stable/11/tools/build/mk/OptionalObsoleteFiles.inc Directory Properties: stable/11/ (props changed) Modified: stable/11/tools/build/mk/OptionalObsoleteFiles.inc == --- stable/11/tools/build/mk/OptionalObsoleteFiles.inc Thu Feb 14 09:48:13 2019(r344116) +++ stable/11/tools/build/mk/OptionalObsoleteFiles.inc Thu Feb 14 09:50:59 2019(r344117) @@ -6537,6 +6537,7 @@ OLD_FILES+=usr/sbin/rpc.ypxfrd OLD_FILES+=usr/sbin/yp_mkdb OLD_FILES+=usr/sbin/ypbind OLD_FILES+=usr/sbin/ypinit +OLD_FILES+=usr/sbin/ypldap OLD_FILES+=usr/sbin/yppoll OLD_FILES+=usr/sbin/yppush OLD_FILES+=usr/sbin/ypserv @@ -6549,6 +6550,7 @@ OLD_FILES+=usr/share/man/man1/ypmatch.1.gz OLD_FILES+=usr/share/man/man1/yppasswd.1.gz OLD_FILES+=usr/share/man/man1/ypwhich.1.gz OLD_FILES+=usr/share/man/man5/netid.5.gz +OLD_FILES+=usr/share/man/man5/ypldap.conf.5.gz OLD_FILES+=usr/share/man/man8/mknetid.8.gz OLD_FILES+=usr/share/man/man8/rpc.yppasswdd.8.gz OLD_FILES+=usr/share/man/man8/rpc.ypxfrd.8.gz @@ -6559,6 +6561,7 @@ OLD_FILES+=usr/share/man/man8/nis.8.gz OLD_FILES+=usr/share/man/man8/yp_mkdb.8.gz OLD_FILES+=usr/share/man/man8/ypbind.8.gz OLD_FILES+=usr/share/man/man8/ypinit.8.gz +OLD_FILES+=usr/share/man/man8/ypldap.8.gz OLD_FILES+=usr/share/man/man8/yppoll.8.gz OLD_FILES+=usr/share/man/man8/yppush.8.gz OLD_FILES+=usr/share/man/man8/ypserv.8.gz @@ -6566,6 +6569,7 @@ OLD_FILES+=usr/share/man/man8/ypset.8.gz OLD_FILES+=usr/share/man/man8/ypxfr.8.gz OLD_FILES+=var/yp/Makefile OLD_FILES+=var/yp/Makefile.dist +OLD_DIRS+=var/yp .endif .if ${MK_NLS} == no ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r344116 - stable/12/tools/build/mk
Author: avos Date: Thu Feb 14 09:48:13 2019 New Revision: 344116 URL: https://svnweb.freebsd.org/changeset/base/344116 Log: MFC r343868: Correct ypldap(8) install path in OptionalObsoleteFiles.inc It's installed to /usr/sbin, not to /usr/bin. While here, add missing manpages and /var/yp directory to the list. Modified: stable/12/tools/build/mk/OptionalObsoleteFiles.inc Directory Properties: stable/12/ (props changed) Modified: stable/12/tools/build/mk/OptionalObsoleteFiles.inc == --- stable/12/tools/build/mk/OptionalObsoleteFiles.inc Thu Feb 14 09:21:19 2019(r344115) +++ stable/12/tools/build/mk/OptionalObsoleteFiles.inc Thu Feb 14 09:48:13 2019(r344116) @@ -6485,7 +6485,6 @@ OLD_FILES+=usr/bin/ypcat OLD_FILES+=usr/bin/ypchfn OLD_FILES+=usr/bin/ypchpass OLD_FILES+=usr/bin/ypchsh -OLD_FILES+=usr/bin/ypldap OLD_FILES+=usr/bin/ypmatch OLD_FILES+=usr/bin/yppasswd OLD_FILES+=usr/bin/ypwhich @@ -6509,6 +6508,7 @@ OLD_FILES+=usr/sbin/rpc.ypxfrd OLD_FILES+=usr/sbin/yp_mkdb OLD_FILES+=usr/sbin/ypbind OLD_FILES+=usr/sbin/ypinit +OLD_FILES+=usr/sbin/ypldap OLD_FILES+=usr/sbin/yppoll OLD_FILES+=usr/sbin/yppush OLD_FILES+=usr/sbin/ypserv @@ -6521,6 +6521,7 @@ OLD_FILES+=usr/share/man/man1/ypmatch.1.gz OLD_FILES+=usr/share/man/man1/yppasswd.1.gz OLD_FILES+=usr/share/man/man1/ypwhich.1.gz OLD_FILES+=usr/share/man/man5/netid.5.gz +OLD_FILES+=usr/share/man/man5/ypldap.conf.5.gz OLD_FILES+=usr/share/man/man8/mknetid.8.gz OLD_FILES+=usr/share/man/man8/rpc.yppasswdd.8.gz OLD_FILES+=usr/share/man/man8/rpc.ypxfrd.8.gz @@ -6531,6 +6532,7 @@ OLD_FILES+=usr/share/man/man8/nis.8.gz OLD_FILES+=usr/share/man/man8/yp_mkdb.8.gz OLD_FILES+=usr/share/man/man8/ypbind.8.gz OLD_FILES+=usr/share/man/man8/ypinit.8.gz +OLD_FILES+=usr/share/man/man8/ypldap.8.gz OLD_FILES+=usr/share/man/man8/yppoll.8.gz OLD_FILES+=usr/share/man/man8/yppush.8.gz OLD_FILES+=usr/share/man/man8/ypserv.8.gz @@ -6538,6 +6540,7 @@ OLD_FILES+=usr/share/man/man8/ypset.8.gz OLD_FILES+=usr/share/man/man8/ypxfr.8.gz OLD_FILES+=var/yp/Makefile OLD_FILES+=var/yp/Makefile.dist +OLD_DIRS+=var/yp .endif .if ${MK_NLS} == no ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r344115 - head
Author: avos Date: Thu Feb 14 09:21:19 2019 New Revision: 344115 URL: https://svnweb.freebsd.org/changeset/base/344115 Log: Add UPDATING entry for IEEE80211_AMPDU_AGE and AH_SUPPORT_AR5416 options removal Notified by: ian Modified: head/UPDATING Modified: head/UPDATING == --- head/UPDATING Thu Feb 14 09:12:19 2019(r344114) +++ head/UPDATING Thu Feb 14 09:21:19 2019(r344115) @@ -38,6 +38,12 @@ NOTE TO PEOPLE WHO THINK THAT FreeBSD 13.x IS SLOW: modules on kernels not having 'device iflib', the iflib.ko module is loaded automatically. +20190125: + The IEEE80211_AMPDU_AGE and AH_SUPPORT_AR5416 kernel configuration + options no longer exist since r343219 and r343427 respectively; + nothing uses them, so they should be just removed from custom + kernel config files. + 20181230: r342635 changes the way efibootmgr(8) works by requiring users to add the -b (bootnum) parameter for commands where the bootnum was previously ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r344114 - stable/12/sbin/sysctl
Author: ygy (doc committer) Date: Thu Feb 14 09:12:19 2019 New Revision: 344114 URL: https://svnweb.freebsd.org/changeset/base/344114 Log: MFC r343930: Remove -R option which was added to sysctl(8) man page per r244106, but it is not implemented. Submitted by: Alfonso Siciliano Reviewed by: 0mp, imp Differential Revision:https://reviews.freebsd.org/D19012 Modified: stable/12/sbin/sysctl/sysctl.8 Directory Properties: stable/12/ (props changed) Modified: stable/12/sbin/sysctl/sysctl.8 == --- stable/12/sbin/sysctl/sysctl.8 Thu Feb 14 00:52:03 2019 (r344113) +++ stable/12/sbin/sysctl/sysctl.8 Thu Feb 14 09:12:19 2019 (r344114) @@ -28,7 +28,7 @@ .\"From: @(#)sysctl.8 8.1 (Berkeley) 6/6/93 .\" $FreeBSD$ .\" -.Dd September 24, 2018 +.Dd February 8, 2019 .Dt SYSCTL 8 .Os .Sh NAME @@ -36,13 +36,13 @@ .Nd get or set kernel state .Sh SYNOPSIS .Nm -.Op Fl bdehiNnoRTtqx +.Op Fl bdehiNnoTtqWx .Op Fl B Ar bufsize .Op Fl f Ar filename .Ar name Ns Op = Ns Ar value Ns Op , Ns Ar value .Ar ... .Nm -.Op Fl bdehNnoRTtqx +.Op Fl bdehNnoTtqWx .Op Fl B Ar bufsize .Fl a .Sh DESCRIPTION ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r343030 - in head/sys: cam conf dev/md dev/nvme fs/fuse fs/nfsclient fs/smbfs kern sys ufs/ffs vm
On Thu, Feb 14, 2019 at 06:56:42PM +1100, Bruce Evans wrote: > I don't understand how pbuf_preallocate() allocates for the other > pbuf pools. When I debugged this for clpbufs, the preallocation was > not used. pbuf types other than clpbufs seem to be unused in my > configurations. I thought that pbufs were used during initialization, > since they end up with a nonzero FREE count, but their only use seems > to be to preallocate them. vnode_pager_generic_getpages() typically not used for UFS on modern systems. Instead the buffer pager is active which does not need pbufs, it uses real buffers coherent with the UFS buffer cache. To get to the actual use of pbufs now you can: - perform clustered buffer io; - use vnode-backed md(4) (this case is still broken if md(4) is loaded as a module); - cause system swapping; - use sendfile(2). ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"