svn commit: r303074 - head/lib/libc/gen
Author: ache Date: Wed Jul 20 07:30:44 2016 New Revision: 303074 URL: https://svnweb.freebsd.org/changeset/base/303074 Log: 1) Per POSIX (and glibc) GLOB_NOCHECK should return original pattern, unmodified, if no matches found. But our original code strips all '\' returning it. Rewrite the code to allow to reconstruct exact the original pattern with backslashes for this case. 2) Prevent to use truncated pattern if MAXPATHLEN exceeded, return GLOB_NOMATCH instead. 3) Fix few end loop conditions filling Char arrays with mbrtowc(), MB_CUR_MAX is unneeded in two places and condition is less by one in other place. 4) Prevent to use truncated filenames match if MAXPATHLEN exceeded, skip such directory entries. 5) Don't end *pathend with L'/' in glob3() if limit is reached, this change will be not visible since error is returned. 6) If error happens in (*readdirfunc)(), do the same GLOB_ABORTED processing as for g_opendir() as POSIX requires. Modified: head/lib/libc/gen/glob.c Modified: head/lib/libc/gen/glob.c == --- head/lib/libc/gen/glob.cWed Jul 20 06:29:26 2016(r303073) +++ head/lib/libc/gen/glob.cWed Jul 20 07:30:44 2016(r303074) @@ -152,6 +152,7 @@ typedef char Char; #defineCHAR(c) ((Char)((c)&M_CHAR)) #defineMETA(c) ((Char)((c)|M_QUOTE)) +#defineUNPROT(c) ((c) & ~M_PROTECT) #defineM_ALL META(L'*') #defineM_END META(L']') #defineM_NOT META(L'!') @@ -159,10 +160,11 @@ typedef char Char; #defineM_RNG META(L'-') #defineM_SET META(L'[') #defineismeta(c) (((c)&M_QUOTE) != 0) +#defineisprot(c) (((c)&M_PROTECT) != 0) static int compare(const void *, const void *); -static int g_Ctoc(const Char *, char *, size_t); +static int g_Ctoc(const Char *, char *, size_t, int); static int g_lstat(Char *, struct stat *, glob_t *); static DIR *g_opendir(Char *, glob_t *); static const Char *g_strchr(const Char *, wchar_t); @@ -176,7 +178,7 @@ static int glob2(Char *, Char *, Char * struct glob_limit *); static int glob3(Char *, Char *, Char *, Char *, Char *, glob_t *, struct glob_limit *); -static int globextend(const Char *, glob_t *, struct glob_limit *); +static int globextend(const Char *, glob_t *, struct glob_limit *, int); static const Char * globtilde(const Char *, Char *, size_t, glob_t *); static int globexp1(const Char *, glob_t *, struct glob_limit *); @@ -197,6 +199,7 @@ glob(const char * __restrict pattern, in mbstate_t mbs; wchar_t wc; size_t clen; + int too_long; patnext = pattern; if (!(flags & GLOB_APPEND)) { @@ -216,24 +219,27 @@ glob(const char * __restrict pattern, in bufnext = patbuf; bufend = bufnext + MAXPATHLEN - 1; + too_long = 1; if (flags & GLOB_NOESCAPE) { memset(&mbs, 0, sizeof(mbs)); - while (bufend - bufnext >= MB_CUR_MAX) { + while (bufnext <= bufend) { clen = mbrtowc(&wc, patnext, MB_LEN_MAX, &mbs); if (clen == (size_t)-1 || clen == (size_t)-2) return (GLOB_NOMATCH); - else if (clen == 0) + else if (clen == 0) { + too_long = 0; break; + } *bufnext++ = wc; patnext += clen; } } else { /* Protect the quoted characters. */ memset(&mbs, 0, sizeof(mbs)); - while (bufend - bufnext >= MB_CUR_MAX) { + while (bufnext <= bufend) { if (*patnext == '\\') { if (*++patnext == '\0') { - *bufnext++ = QUOTE | M_PROTECT; + *bufnext++ = QUOTE; continue; } prot = M_PROTECT; @@ -242,13 +248,16 @@ glob(const char * __restrict pattern, in clen = mbrtowc(&wc, patnext, MB_LEN_MAX, &mbs); if (clen == (size_t)-1 || clen == (size_t)-2) return (GLOB_NOMATCH); - else if (clen == 0) + else if (clen == 0) { + too_long = 0; break; - *bufnext++ = wc | ((wc != DOT && wc != SEP) ? - prot : 0); + } + *bufnext++ = wc | prot; patnext += clen;
Re: svn commit: r303033 - head/share/man/man7
On Tue, 19 Jul 2016, Warner Losh wrote: On Tue, Jul 19, 2016 at 9:01 PM, Jan Beich wrote: Ed Maste writes: +.Ss Predefined Macros +The compiler provides a number of predefined macros. +Some of these provide architecture-specific details and are explained below. +Other macros, including those required by the language standard, are not +included here. [...] +cc -x c -Dm -E /dev/null Typo: -Dm vs. -dM +.It Dv BYTE_ORDER Ta Either Dv BIG_ENDIAN or Dv LITTLE_ENDIAN . Are these really compiler macros? I think, defines them. These are non-compiler-predefined macros. They are namespace pollution in which is only there in the __BSD_VISIBLE case. sys/endian.h defines them (and it implements that by including machine/endian.h in part). $ clang38 -x c -dM -E /dev/null | fgrep ENDIAN #define __BYTE_ORDER__ __ORDER_LITTLE_ENDIAN__ #define __LITTLE_ENDIAN__ 1 #define __ORDER_BIG_ENDIAN__ 4321 #define __ORDER_LITTLE_ENDIAN__ 1234 #define __ORDER_PDP_ENDIAN__ 3412 These are the related compiler definitions in the implementation namespace. Their spelling is quite different and there are no consistency checks between values, but the ABI requires similar values. The values are as different as the spellings in some cases. __LITTLE_ENDIAN__ is 1, but FreeBSD's _LITTLE_ENDIAN is 1234. Actually, the semantics of these is very different too. FreeBSD's _LITTLE_ENDIAN corresponds to __ORDER_LITTLE_ENDIAN. $ gcc5 -x c -dM -E /dev/null | fgrep ENDIAN #define __ORDER_LITTLE_ENDIAN__ 1234 #define __FLOAT_WORD_ORDER__ __ORDER_LITTLE_ENDIAN__ #define __ORDER_PDP_ENDIAN__ 3412 #define __ORDER_BIG_ENDIAN__ 4321 #define __BYTE_ORDER__ __ORDER_LITTLE_ENDIAN__ gcc apparently doesn't even have the confusing __LITTLE_ENDIAN__. This is why they are defined in machine/endian.h. Compilers have been somewhat inconsistent in the past. But they aren't defined in machine/endian.h. Compilers had no trace of them in the past. Even gcc-4.2 in the present /usr/src doesn't have them. The FreeBSD macros should be used since they have been portable in BSD for at last 25 years, with a better spelling which is unsuitable for standards because it is not ugly enough to not be already in use with possibly inconsistent semantics. glibc is different again. In an old version (2.6), it uses __BYTE_ORDER where FreeBSD uses _BYTE_ORDER and compilers use __BYTE_ORDER__. It also defines the FreeBSD pollution BYTE_ORDER in string/endian.h and uses this in dangerous places like netinet/icmp6.h, but in the standard netinet/in.h which doesn't allow this, it is more careful and uses __BYTE_ORDER. The man page might need to be more careful about this. BYTE_ORDER is not a compiler predefine or standard spelling and determing the applicable spelling is time-consuming. 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"
svn commit: r303075 - stable/10/usr.bin/ul
Author: gahr (ports committer) Date: Wed Jul 20 07:33:48 2016 New Revision: 303075 URL: https://svnweb.freebsd.org/changeset/base/303075 Log: MFC r302558: Do not truncate lines longer than 512 chars. PR: 210344 Differential Revision:https://reviews.freebsd.org/D6881 Modified: stable/10/usr.bin/ul/ul.c Directory Properties: stable/10/ (props changed) Modified: stable/10/usr.bin/ul/ul.c == --- stable/10/usr.bin/ul/ul.c Wed Jul 20 07:30:44 2016(r303074) +++ stable/10/usr.bin/ul/ul.c Wed Jul 20 07:33:48 2016(r303075) @@ -78,7 +78,9 @@ structCHAR{ int c_width;/* width or -1 if multi-column char. filler */ } ; -static struct CHARobuf[MAXBUF]; +static struct CHARsobuf[MAXBUF]; /* static output buffer */ +static struct CHAR*obuf = sobuf; +static int buflen = MAXBUF; static int col, maxcol; static int mode; static int halfpos; @@ -155,6 +157,9 @@ main(int argc, char **argv) else filter(f); } + if (obuf != sobuf) { + free(obuf); + } exit(0); } @@ -170,128 +175,148 @@ filter(FILE *f) { wint_t c; int i, w; + int copy; + + copy = 0; + + while ((c = getwc(f)) != WEOF) { + if (col == buflen) { + if (obuf == sobuf) { + obuf = NULL; + copy = 1; + } + obuf = realloc(obuf, sizeof(*obuf) * 2 * buflen); + if (obuf == NULL) { + obuf = sobuf; + break; + } else if (copy) { + memcpy(obuf, sobuf, sizeof(*obuf) * buflen); + copy = 0; + } + bzero((char *)(obuf + buflen), sizeof(*obuf) * buflen); + buflen *= 2; + } + switch(c) { + case '\b': + if (col > 0) + col--; + continue; - while ((c = getwc(f)) != WEOF && col < MAXBUF) switch(c) { + case '\t': + col = (col+8) & ~07; + if (col > maxcol) + maxcol = col; + continue; - case '\b': - if (col > 0) - col--; - continue; - - case '\t': - col = (col+8) & ~07; - if (col > maxcol) - maxcol = col; - continue; - - case '\r': - col = 0; - continue; - - case SO: - mode |= ALTSET; - continue; - - case SI: - mode &= ~ALTSET; - continue; - - case IESC: - switch (c = getwc(f)) { - - case HREV: - if (halfpos == 0) { - mode |= SUPERSC; - halfpos--; - } else if (halfpos > 0) { - mode &= ~SUBSC; - halfpos--; - } else { - halfpos = 0; - reverse(); - } + case '\r': + col = 0; continue; - case HFWD: - if (halfpos == 0) { - mode |= SUBSC; - halfpos++; - } else if (halfpos < 0) { - mode &= ~SUPERSC; - halfpos++; - } else { - halfpos = 0; - fwd(); - } + case SO: + mode |= ALTSET; continue; - case FREV: - reverse(); + case SI: + mode &= ~ALTSET; continue; - default: - errx(1, "unknown escape sequence in input: %o, %o", IESC, c); - } - continue; + case IESC: + switch (c = getwc(f)) { - case '_': - if (obuf[col].c_char || obuf[col].c_width < 0) { - while (col > 0 && obuf[col].c_width < 0) - col--; - w = obuf[col].c_width; - for (i = 0; i < w; i++) - obuf[col++].c_mode |= UNDERL | mode; + case HREV: +
Re: svn commit: r303043 - in head: share/man/man4 sys/dev/vt sys/dev/vt/hw/fb
On 20/07/2016 5:20 AM, Conrad E. Meyer wrote: > Author: cem > Date: Tue Jul 19 19:20:47 2016 > New Revision: 303043 > URL: https://svnweb.freebsd.org/changeset/base/303043 > > Log: > Increase vt(4) framebuffer maximum size > > And rename "DEFAULT" constants to the more accurate "MAX." > > PR: 210382 > Submitted by: Felix > Reviewed by:wblock, cem > Tested by: Dave Cottlehuber Can MFC? There are high hopes for this coming out in 11.0 / next 10.x re is cc'd on the original bugzilla issue > Modified: > head/share/man/man4/vt.4 > head/sys/dev/vt/hw/fb/vt_fb.c > head/sys/dev/vt/vt.h > head/sys/dev/vt/vt_core.c > > Modified: head/share/man/man4/vt.4 > == > --- head/share/man/man4/vt.4 Tue Jul 19 19:19:03 2016(r303042) > +++ head/share/man/man4/vt.4 Tue Jul 19 19:20:47 2016(r303043) > @@ -24,7 +24,7 @@ > .\" > .\" $FreeBSD$ > .\" > -.Dd June 21, 2016 > +.Dd July 19, 2016 > .Dt "VIRTUAL TERMINALS" 4 > .Os > .Sh NAME > @@ -36,8 +36,8 @@ > .Cd "options VT_MAXWINDOWS=N" > .Cd "options VT_ALT_TO_ESC_HACK=1" > .Cd "options VT_TWOBUTTON_MOUSE" > -.Cd "options VT_FB_DEFAULT_WIDTH=X" > -.Cd "options VT_FB_DEFAULT_HEIGHT=Y" > +.Cd "options VT_FB_MAX_WIDTH=X" > +.Cd "options VT_FB_MAX_HEIGHT=Y" > .Cd "options SC_NO_CUTPASTE" > .Cd "device vt" > .Pp > > Modified: head/sys/dev/vt/hw/fb/vt_fb.c > == > --- head/sys/dev/vt/hw/fb/vt_fb.c Tue Jul 19 19:19:03 2016 > (r303042) > +++ head/sys/dev/vt/hw/fb/vt_fb.c Tue Jul 19 19:20:47 2016 > (r303043) > @@ -416,10 +416,10 @@ vt_fb_init(struct vt_device *vd) > int err; > > info = vd->vd_softc; > - vd->vd_height = MIN(VT_FB_DEFAULT_HEIGHT, info->fb_height); > + vd->vd_height = MIN(VT_FB_MAX_HEIGHT, info->fb_height); > margin = (info->fb_height - vd->vd_height) >> 1; > vd->vd_transpose = margin * info->fb_stride; > - vd->vd_width = MIN(VT_FB_DEFAULT_WIDTH, info->fb_width); > + vd->vd_width = MIN(VT_FB_MAX_WIDTH, info->fb_width); > margin = (info->fb_width - vd->vd_width) >> 1; > vd->vd_transpose += margin * (info->fb_bpp / NBBY); > vd->vd_video_dev = info->fb_video_dev; > > Modified: head/sys/dev/vt/vt.h > == > --- head/sys/dev/vt/vt.h Tue Jul 19 19:19:03 2016(r303042) > +++ head/sys/dev/vt/vt.h Tue Jul 19 19:20:47 2016(r303043) > @@ -377,11 +377,11 @@ void vt_upgrade(struct vt_device *vd); > #define PIXEL_WIDTH(w) ((w) / 8) > #define PIXEL_HEIGHT(h) ((h) / 16) > > -#ifndef VT_FB_DEFAULT_WIDTH > -#define VT_FB_DEFAULT_WIDTH 2048 > +#ifndef VT_FB_MAX_WIDTH > +#define VT_FB_MAX_WIDTH 4096 > #endif > -#ifndef VT_FB_DEFAULT_HEIGHT > -#define VT_FB_DEFAULT_HEIGHT1200 > +#ifndef VT_FB_MAX_HEIGHT > +#define VT_FB_MAX_HEIGHT2400 > #endif > > /* name argument is not used yet. */ > > Modified: head/sys/dev/vt/vt_core.c > == > --- head/sys/dev/vt/vt_core.c Tue Jul 19 19:19:03 2016(r303042) > +++ head/sys/dev/vt/vt_core.c Tue Jul 19 19:20:47 2016(r303043) > @@ -181,8 +181,8 @@ static void vt_resume_handler(void *priv > > SET_DECLARE(vt_drv_set, struct vt_driver); > > -#define _VTDEFH MAX(100, PIXEL_HEIGHT(VT_FB_DEFAULT_HEIGHT)) > -#define _VTDEFW MAX(200, PIXEL_WIDTH(VT_FB_DEFAULT_WIDTH)) > +#define _VTDEFH MAX(100, PIXEL_HEIGHT(VT_FB_MAX_HEIGHT)) > +#define _VTDEFW MAX(200, PIXEL_WIDTH(VT_FB_MAX_WIDTH)) > > struct terminal vt_consterm; > static struct vt_window vt_conswindow; > ___ > svn-src-h...@freebsd.org mailing list > https://lists.freebsd.org/mailman/listinfo/svn-src-head > To unsubscribe, send any mail to "svn-src-head-unsubscr...@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: r303076 - head/sys/dev/fb
Author: royger Date: Wed Jul 20 09:29:39 2016 New Revision: 303076 URL: https://svnweb.freebsd.org/changeset/base/303076 Log: vesa: fix panic on suspend Fix the following panic seen when migrating a FreeBSD guest on Xen: panic: mtx_lock() of destroyed mutex @ /usr/src/sys/dev/fb/vesa.c:541 cpuid = 0 KDB: stack backtrace: db_trace_self_wrapper() at db_trace_self_wrapper+0x2b/frame 0xfe001d2fa4f0 vpanic() at vpanic+0x182/frame 0xfe001d2fa570 kassert_panic() at kassert_panic+0x126/frame 0xfe001d2fa5e0 __mtx_lock_flags() at __mtx_lock_flags+0x15b/frame 0xfe001d2fa630 vesa_bios_save_restore() at vesa_bios_save_restore+0x78/frame 0xfe001d2fa680 vga_suspend() at vga_suspend+0xa3/frame 0xfe001d2fa6b0 isavga_suspend() at isavga_suspend+0x1d/frame 0xfe001d2fa6d0 bus_generic_suspend_child() at bus_generic_suspend_child+0x44/frame [...] This is caused because vga_sub_configure (which is called if the VGA adapter is attached after VESA tried to initialize), points to vesa_configure, which doesn't initialize the VESA mutex. In order to fix it, make sure vga_sub_configure points to vesa_load, so that all the needed vesa components are properly initialized. Sponsored by: Citrix Systems R&D MFC after:3 days PR: 209203 Reviewed by: dumbbell Differential revision:https://reviews.freebsd.org/D7196 Modified: head/sys/dev/fb/vesa.c Modified: head/sys/dev/fb/vesa.c == --- head/sys/dev/fb/vesa.c Wed Jul 20 07:33:48 2016(r303075) +++ head/sys/dev/fb/vesa.c Wed Jul 20 09:29:39 2016(r303076) @@ -134,6 +134,7 @@ static vi_fill_rect_t vesa_fill_rect; static vi_bitblt_t vesa_bitblt; static vi_diag_t vesa_diag; static int vesa_bios_info(int level); +static int vesa_late_load(int flags); static video_switch_t vesavidsw = { vesa_probe, @@ -1141,7 +1142,7 @@ vesa_configure(int flags) * initialization for now and try again later. */ if (adp == NULL) { - vga_sub_configure = vesa_configure; + vga_sub_configure = vesa_late_load; return (ENODEV); } @@ -1909,6 +1910,17 @@ vesa_bios_info(int level) static int vesa_load(void) { + + return (vesa_late_load(0)); +} + +/* + * To be called from the vga_sub_configure hook in case the VGA adapter is + * not found when VESA is loaded. + */ +static int +vesa_late_load(int flags) +{ int error; if (vesa_init_done) @@ -1918,7 +1930,7 @@ vesa_load(void) /* locate a VGA adapter */ vesa_adp = NULL; - error = vesa_configure(0); + error = vesa_configure(flags); if (error == 0) vesa_bios_info(bootverbose); ___ 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: r303077 - vendor-sys/illumos/dist/uts/common/fs/zfs
Author: avg Date: Wed Jul 20 09:47:35 2016 New Revision: 303077 URL: https://svnweb.freebsd.org/changeset/base/303077 Log: 7072 zfs fails to expand if lun added when os is in shutdown state illumos/illumos-gate@c39a2aae1e2c439d156021edfc20910dad7f9891 https://github.com/illumos/illumos-gate/commit/c39a2aae1e2c439d156021edfc20910dad7f9891 https://www.illumos.org/issues/7072 upstream: 38733 zfs fails to expand if lun added when os is in shutdown state DLPX-36910 spares and caches should not display expandable space DLPX-39262 vdev_disk_open spam zfs_dbgmsg buffer Reviewed by: Igor Kozhukhov Reviewed by: Dan Kimmel Reviewed by: Matthew Ahrens Reviewed by: Prakash Surya Reviewed by: Alex Reece Approved by: Dan McDonald Author: George Wilson Modified: vendor-sys/illumos/dist/uts/common/fs/zfs/metaslab.c vendor-sys/illumos/dist/uts/common/fs/zfs/vdev.c vendor-sys/illumos/dist/uts/common/fs/zfs/vdev_disk.c Modified: vendor-sys/illumos/dist/uts/common/fs/zfs/metaslab.c == --- vendor-sys/illumos/dist/uts/common/fs/zfs/metaslab.cWed Jul 20 09:29:39 2016(r303076) +++ vendor-sys/illumos/dist/uts/common/fs/zfs/metaslab.cWed Jul 20 09:47:35 2016(r303077) @@ -381,7 +381,13 @@ metaslab_class_expandable_space(metaslab continue; } - space += tvd->vdev_max_asize - tvd->vdev_asize; + /* +* Calculate if we have enough space to add additional +* metaslabs. We report the expandable space in terms +* of the metaslab size since that's the unit of expansion. +*/ + space += P2ALIGN(tvd->vdev_max_asize - tvd->vdev_asize, + 1ULL << tvd->vdev_ms_shift); } spa_config_exit(mc->mc_spa, SCL_VDEV, FTAG); return (space); Modified: vendor-sys/illumos/dist/uts/common/fs/zfs/vdev.c == --- vendor-sys/illumos/dist/uts/common/fs/zfs/vdev.cWed Jul 20 09:29:39 2016(r303076) +++ vendor-sys/illumos/dist/uts/common/fs/zfs/vdev.cWed Jul 20 09:47:35 2016(r303077) @@ -2731,6 +2731,7 @@ vdev_get_stats(vdev_t *vd, vdev_stat_t * { spa_t *spa = vd->vdev_spa; vdev_t *rvd = spa->spa_root_vdev; + vdev_t *tvd = vd->vdev_top; ASSERT(spa_config_held(spa, SCL_ALL, RW_READER) != 0); @@ -2741,7 +2742,15 @@ vdev_get_stats(vdev_t *vd, vdev_stat_t * vs->vs_rsize = vdev_get_min_asize(vd); if (vd->vdev_ops->vdev_op_leaf) vs->vs_rsize += VDEV_LABEL_START_SIZE + VDEV_LABEL_END_SIZE; - vs->vs_esize = vd->vdev_max_asize - vd->vdev_asize; + /* +* Report expandable space on top-level, non-auxillary devices only. +* The expandable space is reported in terms of metaslab sized units +* since that determines how much space the pool can expand. +*/ + if (vd->vdev_aux == NULL && tvd != NULL) { + vs->vs_esize = P2ALIGN(vd->vdev_max_asize - vd->vdev_asize, + 1ULL << tvd->vdev_ms_shift); + } if (vd->vdev_aux == NULL && vd == vd->vdev_top && !vd->vdev_ishole) { vs->vs_fragmentation = vd->vdev_mg->mg_fragmentation; } Modified: vendor-sys/illumos/dist/uts/common/fs/zfs/vdev_disk.c == --- vendor-sys/illumos/dist/uts/common/fs/zfs/vdev_disk.c Wed Jul 20 09:29:39 2016(r303076) +++ vendor-sys/illumos/dist/uts/common/fs/zfs/vdev_disk.c Wed Jul 20 09:47:35 2016(r303077) @@ -241,34 +241,6 @@ vdev_disk_rele(vdev_t *vd) } } -static uint64_t -vdev_disk_get_space(vdev_t *vd, uint64_t capacity, uint_t blksz) -{ - ASSERT(vd->vdev_wholedisk); - - vdev_disk_t *dvd = vd->vdev_tsd; - dk_efi_t dk_ioc; - efi_gpt_t *efi; - uint64_t avail_space = 0; - int efisize = EFI_LABEL_SIZE * 2; - - dk_ioc.dki_data = kmem_alloc(efisize, KM_SLEEP); - dk_ioc.dki_lba = 1; - dk_ioc.dki_length = efisize; - dk_ioc.dki_data_64 = (uint64_t)(uintptr_t)dk_ioc.dki_data; - efi = dk_ioc.dki_data; - - if (ldi_ioctl(dvd->vd_lh, DKIOCGETEFI, (intptr_t)&dk_ioc, - FKIOCTL, kcred, NULL) == 0) { - uint64_t efi_altern_lba = LE_64(efi->efi_gpt_AlternateLBA); - - if (capacity > efi_altern_lba) - avail_space = (capacity - efi_altern_lba) * blksz; - } - kmem_free(dk_ioc.dki_data, efisize); - return (avail_space); -} - /* * We want to be loud in DEBUG kernels when DKIOCGMEDIAINFOEXT fails, or when * even a fallback to DKIOCGMEDIAINFO fails. @@ -559,10 +531,7 @@ skip_open: * Adjust max_psize upward accordingly sinc
svn commit: r303078 - vendor-sys/illumos/dist/uts/common/fs/zfs
Author: avg Date: Wed Jul 20 09:49:09 2016 New Revision: 303078 URL: https://svnweb.freebsd.org/changeset/base/303078 Log: 7086 ztest attempts dva_get_dsize_sync on an embedded blockpointer illumos/illumos-gate@926549256b71acd595f69b236779ff6b78fa08ef https://github.com/illumos/illumos-gate/commit/926549256b71acd595f69b236779ff6b78fa08ef https://www.illumos.org/issues/7086 In dbuf_dirty(), we need to grab the dn_struct_rwlock before looking at the db_blkptr, to prevent it from being changed by syncing context. Otherwise we may see that ztest got a segfault from this stack: libzpool.so.1`dva_get_dsize_sync+0x98(872f000, b32b240, fed7811b, 0, b4cda20, 0) libzpool.so.1`bp_get_dsize+0x60(872f000, b32b240, 0, 97cb780, 9d4c1a8, 0) libzpool.so.1`dbuf_dirty+0x9b3(ce0a100, 97cb780, 9, fecd2530) libzpool.so.1`dmu_buf_will_dirty+0xc3(ce0a100, 97cb780, ea293d6c, 1) libzpool.so.1`zap_lockdir+0x1a0(8aaa3c0, 1, 0, 97cb780, 1, 1) libzpool.so.1`zap_remove_norm+0x30(8aaa3c0, 1, 0, 8728b10, 0, 97cb780) libzpool.so.1`zap_remove+0x29(8aaa3c0, 1, 0, 8728b10, 97cb780, a) ztest_replay_remove+0x225(ea294588, 8728ae8, 0, 3801, 0, 0) ztest_remove+0x9f(ea294588, ea293f50, 4, 3) ztest_object_init+0x78(ea294588, ea293f50, 4e0, 1) ztest_dmu_object_alloc_free+0x71(ea294588, 13) ztest_dmu_objset_create_destroy+0x224(80cef08, 13, 0, 805d36c, 9017ad44, 0) ztest_execute+0x89(a, 807c720, 13, 0) ztest_thread+0xea(13, 0, 0, 0) libc.so.1`_thrp_setup+0x88(f0983240) libc.so.1`_lwp_start(f0983240, 0, 0, 0, 0, 0) Looking into it a bit, we see that this is an embedded blockpointer, so BP_GET_NDVAS should have returned 0: b32b240::blkptr EMBEDDED [L0 ZAP_OTHER] et=0 LZ4 size=200L/4aP birth=80L Instead, it looks like another thread is modifying this blockpointer: b32b240::ugrep | ::whatis f47a0e0c is in [ stack tid=0x19f ] ebd6ec40 is in [ stack tid=0x226 ] ea293bd0 is in [ stack tid=0x244 ] ea293be4 is in [ stack tid=0x244 ] Reviewed by: Prakash Surya Reviewed by: George Wilson Approved by: Robert Mustacchi Author: Matthew Ahrens Modified: vendor-sys/illumos/dist/uts/common/fs/zfs/dbuf.c Modified: vendor-sys/illumos/dist/uts/common/fs/zfs/dbuf.c == --- vendor-sys/illumos/dist/uts/common/fs/zfs/dbuf.cWed Jul 20 09:47:35 2016(r303077) +++ vendor-sys/illumos/dist/uts/common/fs/zfs/dbuf.cWed Jul 20 09:49:09 2016(r303078) @@ -1662,7 +1662,20 @@ dbuf_dirty(dmu_buf_impl_t *db, dmu_tx_t dnode_setdirty(dn, tx); DB_DNODE_EXIT(db); return (dr); - } else if (do_free_accounting) { + } + + /* +* The dn_struct_rwlock prevents db_blkptr from changing +* due to a write from syncing context completing +* while we are running, so we want to acquire it before +* looking at db_blkptr. +*/ + if (!RW_WRITE_HELD(&dn->dn_struct_rwlock)) { + rw_enter(&dn->dn_struct_rwlock, RW_READER); + drop_struct_lock = TRUE; + } + + if (do_free_accounting) { blkptr_t *bp = db->db_blkptr; int64_t willfree = (bp && !BP_IS_HOLE(bp)) ? bp_get_dsize(os->os_spa, bp) : db->db.db_size; @@ -1678,11 +1691,6 @@ dbuf_dirty(dmu_buf_impl_t *db, dmu_tx_t dnode_willuse_space(dn, -willfree, tx); } - if (!RW_WRITE_HELD(&dn->dn_struct_rwlock)) { - rw_enter(&dn->dn_struct_rwlock, RW_READER); - drop_struct_lock = TRUE; - } - if (db->db_level == 0) { dnode_new_blkid(dn, db->db_blkid, tx, drop_struct_lock); ASSERT(dn->dn_maxblkid >= db->db_blkid); ___ 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: r303079 - vendor/illumos/dist/cmd/ztest
Author: avg Date: Wed Jul 20 09:51:33 2016 New Revision: 303079 URL: https://svnweb.freebsd.org/changeset/base/303079 Log: 7147 ztest: ztest_ddt_repair fails with ztest_pattern_match assertion illumos/illumos-gate@aab80726335c76a7cae32c7300890248d73a51e3 https://github.com/illumos/illumos-gate/commit/aab80726335c76a7cae32c7300890248d73a51e3 https://www.illumos.org/issues/7147 Here's the dbuf we're currently reading: 966f200::dbuf addr object lvl blkid holds os 966f200 4 0 0 1 ztest/ds_3 966f200::print dmu_buf_t db_data db_data = 0x9ae0400 0x9ae0400/10J 0x9ae0400: c1c7ced932020d c1c7ced932020d c1c7ced932020d c1c7ced932020d c1c7ced932020d c1c7ced932020d c1c7ced932020d c1c7ced932020d c1c7ced932020d c1c7ced932020d The pattern we're expecting is actually this: a34ae10b5f2db2. If we attempt to read the block on disk we find that it has matches what ztest_ddt_repair() would have written: ~c1c7ced932020d=J ff3e383126cdfdf2 966f200::print dmu_buf_impl_t db_blkptr | ::blkptr DVA0=<0:71d3c00:800> [L0 UINT64_OTHER] SHA256 OFF LE contiguous dedup single size=400L/400P birth=55L/55P fill=1 cksum=18486450d3ce8c6d:75a72f4bbf117b0f:2d3a226314eb5650:2eb0fd68648b1af0 1. zdb -U /rpool/tmp/zpool.cache -R ztest 0:71d3c00:800 | head Found vdev type: mirror 0:71d3c00:800 0 1 2 3 4 5 6 7 8 9 a b c d e f 0123456789abcdef 00: ff3e383126cdfdf2 ff3e383126cdfdf2 ...&18>&18>. 10: ff3e383126cdfdf2 ff3e383126cdfdf2 ...&18>&18>. 20: ff3e383126cdfdf2 ff3e383126cdfdf2 ...&18>&18>. 30: ff3e383126cdfdf2 ff3e383126cdfdf2 ...&18>&18>. 40: ff3e383126cdfdf2 ff3e383126cdfdf2 ...&18>&18>. 50: ff3e383126cdfdf2 ff3e383126cdfdf2 ...&18>&18>. Reviewed by: Matthew Ahrens Reviewed by: Prakash Surya Approved by: Robert Mustacchi Author: George Wilson Modified: vendor/illumos/dist/cmd/ztest/ztest.c Modified: vendor/illumos/dist/cmd/ztest/ztest.c == --- vendor/illumos/dist/cmd/ztest/ztest.c Wed Jul 20 09:49:09 2016 (r303078) +++ vendor/illumos/dist/cmd/ztest/ztest.c Wed Jul 20 09:51:33 2016 (r303079) @@ -20,7 +20,7 @@ */ /* * Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved. - * Copyright (c) 2011, 2015 by Delphix. All rights reserved. + * Copyright (c) 2011, 2016 by Delphix. All rights reserved. * Copyright 2011 Nexenta Systems, Inc. All rights reserved. * Copyright (c) 2013 Steven Hartland. All rights reserved. * Copyright (c) 2014 Integros [integros.com] @@ -5021,9 +5021,14 @@ ztest_ddt_repair(ztest_ds_t *zd, uint64_ return; } + dmu_objset_stats_t dds; + dsl_pool_config_enter(dmu_objset_pool(os), FTAG); + dmu_objset_fast_stat(os, &dds); + dsl_pool_config_exit(dmu_objset_pool(os), FTAG); + object = od[0].od_object; blocksize = od[0].od_blocksize; - pattern = zs->zs_guid ^ dmu_objset_fsid_guid(os); + pattern = zs->zs_guid ^ dds.dds_guid; ASSERT(object != 0); @@ -5625,9 +5630,13 @@ ztest_run(ztest_shared_t *zs) metaslab_preload_limit = ztest_random(20) + 1; ztest_spa = spa; + dmu_objset_stats_t dds; VERIFY0(dmu_objset_own(ztest_opts.zo_pool, DMU_OST_ANY, B_TRUE, FTAG, &os)); - zs->zs_guid = dmu_objset_fsid_guid(os); + dsl_pool_config_enter(dmu_objset_pool(os), FTAG); + dmu_objset_fast_stat(os, &dds); + dsl_pool_config_exit(dmu_objset_pool(os), FTAG); + zs->zs_guid = dds.dds_guid; dmu_objset_disown(os, FTAG); spa->spa_dedup_ditto = 2 * ZIO_DEDUPDITTO_MIN; ___ 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: r303080 - vendor/illumos/dist/cmd/ztest
Author: avg Date: Wed Jul 20 09:53:46 2016 New Revision: 303080 URL: https://svnweb.freebsd.org/changeset/base/303080 Log: 6451 ztest fails due to checksum errors illumos/illumos-gate@f9eb9fdf196b6ed476e4ffc69cecd8b0da3cb7e7 https://github.com/illumos/illumos-gate/commit/f9eb9fdf196b6ed476e4ffc69cecd8b0da3cb7e7 https://www.illumos.org/issues/6451 Sometimes ztest fails because zdb detects checksum errors. e.g.: Traversing all blocks to verify checksums and verify nothing leaked ... zdb_blkptr_cb: Got error 50 reading <71, 47, 0, 8000160> DVA0=<0:1cc2000: 18> [L0 other uint64[]] sha256 uncompressed LE contiguou s unique single size=10L/10P birth=271L/271P fill=1 cksum=c5a3e27d1ed0f894:843bca3a5473c4bf:f76a19b6830a2e4:91292591613a12bf -- skipping zdb_blkptr_cb: Got error 50 reading <71, 47, 0, 80180> DVA0=<0:ce16800: 18> [L0 other uint64[]] sha256 uncompressed LE contigu ous unique single size=10L/10P birth=840L/840P fill=1 cksum=5d018f3d061e17f3:6d1584784587bf63:2805a74a0ce37369:ba68a214806c7e75 -- skipping zdb_blkptr_cb: Got error 50 reading <71, 47, 0, 100360> DVA0=<0:10d37400: 18> [L0 other uint64[]] sha256 uncompressed LE conti guous unique single size=10L/10P birth=904L/904P fill=1 cksum=fa1e11d4138bd14b:86c9488c73e3:f31e43c72e72e46b:e3446472d1174d ba -- skipping zdb_blkptr_cb: Got error 50 reading <71, 47, 0, 40002c0> DVA0=<0:127ef400: 18> [L0 other uint64[]] sha256 uncompressed LE cont iguous dedup single size=10L/10P birth=549L/549P fill=1 cksum=30e14955ebf13522:66dc2ff8067e6810:4607e750abb9d3b3:6582b8af909fcb 58 -- skipping zdb_blkptr_cb: Got error 50 reading <657, 5, 0, 1c0> DVA0=<0:1a180400:18> [L0 other uint64[]] fletcher4 uncompressed LE contiguou s unique single size=10L/10P birth=1091L/1091P fill=1 cksum=a6cf1e50: 29b3bd01c57e5:36779b914035db9a:db61cdcf6bec56f0 -- skippin g The problem is that ztest_fault_inject() can inject multiple faults into the same block. It is designed such that it can inject errors on all leafs of a RAID-Z or mirror, but for a given range of offsets, it will only inject errors Reviewed by: George Wilson Reviewed by: Prakash Surya Reviewed by: Jorgen Lundman Approved by: Dan McDonald Author: Matthew Ahrens Modified: vendor/illumos/dist/cmd/ztest/ztest.c Modified: vendor/illumos/dist/cmd/ztest/ztest.c == --- vendor/illumos/dist/cmd/ztest/ztest.c Wed Jul 20 09:51:33 2016 (r303079) +++ vendor/illumos/dist/cmd/ztest/ztest.c Wed Jul 20 09:53:46 2016 (r303080) @@ -4792,7 +4792,7 @@ ztest_fault_inject(ztest_ds_t *zd, uint6 char path0[MAXPATHLEN]; char pathrand[MAXPATHLEN]; size_t fsize; - int bshift = SPA_OLD_MAXBLOCKSHIFT + 2; /* don't scrog all labels */ + int bshift = SPA_MAXBLOCKSHIFT + 2; /* don't scrog all labels */ int iters = 1000; int maxfaults; int mirror_save; @@ -4953,6 +4953,31 @@ ztest_fault_inject(ztest_ds_t *zd, uint6 fsize = lseek(fd, 0, SEEK_END); while (--iters != 0) { + /* +* The offset must be chosen carefully to ensure that +* we do not inject a given logical block with errors +* on two different leaf devices, because ZFS can not +* tolerate that (if maxfaults==1). +* +* We divide each leaf into chunks of size +* (# leaves * SPA_MAXBLOCKSIZE * 4). Within each chunk +* there is a series of ranges to which we can inject errors. +* Each range can accept errors on only a single leaf vdev. +* The error injection ranges are separated by ranges +* which we will not inject errors on any device (DMZs). +* Each DMZ must be large enough such that a single block +* can not straddle it, so that a single block can not be +* a target in two different injection ranges (on different +* leaf vdevs). +* +* For example, with 3 leaves, each chunk looks like: +*0 to 32M: injection range for leaf 0 +* 32M to 64M: DMZ - no injection allowed +* 64M to 96M: injection range for leaf 1 +* 96M to 128M: DMZ - no injection allowed +* 128M to 160M: injection range for leaf 2 +* 160M to 192M: DMZ - no injection allowed +*/ offset = ztest_random(fsize / (leaves << bshift)) * (leaves << bshift) + (leaf << bshift) + (ztest_random(1ULL << (bshift - 1)) & -8ULL); ___ svn-src-a
svn commit: r303081 - vendor/illumos/dist/cmd/ztest
Author: avg Date: Wed Jul 20 09:54:18 2016 New Revision: 303081 URL: https://svnweb.freebsd.org/changeset/base/303081 Log: 7163 ztest failures due to excess error injection illumos/illumos-gate@f34284d835bc555f987c1310df46c034c3101155 https://github.com/illumos/illumos-gate/commit/f34284d835bc555f987c1310df46c034c3101155 https://www.illumos.org/issues/7163 Running zloop from zfs-precommit hit this assertion: *panicstr/s 0xfd7fd7419370: assertion failed for thread 0xfd7fe29ed240, thread-id 577: parent != NULL, file ../../../uts/common/fs/zfs/dbuf.c, line 1827 $c libc.so.1`_lwp_kill+0xa() libc.so.1`_assfail+0x182(fd7ffb1c29fa, fd7ffb1cc028, 723) libc.so.1`assfail+0x19(fd7ffb1c29fa, fd7ffb1cc028, 723) libzpool.so.1`dbuf_dirty+0xc69(10e3bc10, 3601700) libzpool.so.1`dbuf_dirty+0x61e(10c73640, 3601700) libzpool.so.1`dbuf_dirty+0x61e(10e28280, 3601700) libzpool.so.1`dmu_buf_will_fill+0x64(10e28280, 3601700) libzpool.so.1`dmu_write+0x1b6(2c7e640, d, 40002e00, 200, 3717b40, 3601700) ztest_replay_write+0x568(4950d0, 3717a80, 0) ztest_write+0x125(4950d0, d, 40002e00, 200, 413f000) ztest_io+0x1bb(4950d0, d, 40002e00) ztest_dmu_write_parallel+0xaa(4950d0, 6) ztest_execute+0x83(1, 420c98, 6) ztest_thread+0xf4(6) libc.so.1`_thrp_setup+0x8a(fd7fe29ed240) libc.so.1`_lwp_start() This is another manifestation of ECKSUM in ztest: The lowest level ancestor that’s in memory is the L8 (topmost). The L7 ancestor is blkid 0x10: ::dbufs -O 0x2c7e640 -o d -l 7 |::dbuf addr object lvl blkid holds os 600be50 d 7 4 1 ztest/ds_6 719d880 d 7 0 4 ztest/ds_6 Reviewed by: George Wilson Reviewed by: Paul Dagnelie Approved by: Robert Mustacchi Author: Matthew Ahrens Modified: vendor/illumos/dist/cmd/ztest/ztest.c Modified: vendor/illumos/dist/cmd/ztest/ztest.c == --- vendor/illumos/dist/cmd/ztest/ztest.c Wed Jul 20 09:53:46 2016 (r303080) +++ vendor/illumos/dist/cmd/ztest/ztest.c Wed Jul 20 09:54:18 2016 (r303081) @@ -4792,7 +4792,7 @@ ztest_fault_inject(ztest_ds_t *zd, uint6 char path0[MAXPATHLEN]; char pathrand[MAXPATHLEN]; size_t fsize; - int bshift = SPA_MAXBLOCKSHIFT + 2; /* don't scrog all labels */ + int bshift = SPA_MAXBLOCKSHIFT + 2; int iters = 1000; int maxfaults; int mirror_save; @@ -4982,7 +4982,29 @@ ztest_fault_inject(ztest_ds_t *zd, uint6 (leaves << bshift) + (leaf << bshift) + (ztest_random(1ULL << (bshift - 1)) & -8ULL); - if (offset >= fsize) + /* +* Only allow damage to the labels at one end of the vdev. +* +* If all labels are damaged, the device will be totally +* inaccessible, which will result in loss of data, +* because we also damage (parts of) the other side of +* the mirror/raidz. +* +* Additionally, we will always have both an even and an +* odd label, so that we can handle crashes in the +* middle of vdev_config_sync(). +*/ + if ((leaf & 1) == 0 && offset < VDEV_LABEL_START_SIZE) + continue; + + /* +* The two end labels are stored at the "end" of the disk, but +* the end of the disk (vdev_psize) is aligned to +* sizeof (vdev_label_t). +*/ + uint64_t psize = P2ALIGN(fsize, sizeof (vdev_label_t)); + if ((leaf & 1) == 1 && + offset + sizeof (bad) > psize - VDEV_LABEL_END_SIZE) continue; VERIFY(mutex_lock(&ztest_vdev_lock) == 0); ___ 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: r303082 - vendor/illumos/dist/cmd/zdb
Author: avg Date: Wed Jul 20 09:57:16 2016 New Revision: 303082 URL: https://svnweb.freebsd.org/changeset/base/303082 Log: 6391 Override default SPA config location via environment illumos/illumos-gate@ae24175b2b25e9fb4bfd9ac0648b57e0735b6651 https://github.com/illumos/illumos-gate/commit/ae24175b2b25e9fb4bfd9ac0648b57e0735b6651 https://www.illumos.org/issues/6391 When using zdb with non-default SPA config file it is not convenient to add -U all the time. This commit introduces support for setting/overriding SPA config location via environment variable 'SPA_CONFIG_PATH'. If -U flag is specified in the command line it will override any other value as usual. https://github.com/zfsonlinux/zfs/commit/64d7b6cf75e52a4698d9bdec61745573c39d2e5a Reviewed by: Brian Behlendorf Reviewed by: Richard Yao Reviewed by: Matthew Ahrens Reviewed by: Will Andrews Reviewed by: George Wilson Approved by: Robert Mustacchi Author: Cyril Plisko Modified: vendor/illumos/dist/cmd/zdb/zdb.c Modified: vendor/illumos/dist/cmd/zdb/zdb.c == --- vendor/illumos/dist/cmd/zdb/zdb.c Wed Jul 20 09:54:18 2016 (r303081) +++ vendor/illumos/dist/cmd/zdb/zdb.c Wed Jul 20 09:57:16 2016 (r303082) @@ -3529,12 +3529,22 @@ main(int argc, char **argv) nvlist_t *policy = NULL; uint64_t max_txg = UINT64_MAX; int rewind = ZPOOL_NEVER_REWIND; + char *spa_config_path_env; (void) setrlimit(RLIMIT_NOFILE, &rl); (void) enable_extended_FILE_stdio(-1, -1); dprintf_setup(&argc, argv); + /* +* If there is an environment variable SPA_CONFIG_PATH it overrides +* default spa_config_path setting. If -U flag is specified it will +* override this environment variable settings once again. +*/ + spa_config_path_env = getenv("SPA_CONFIG_PATH"); + if (spa_config_path_env != NULL) + spa_config_path = spa_config_path_env; + while ((c = getopt(argc, argv, "bcdhilmMI:suCDRSAFLXx:evp:t:U:P")) != -1) { switch (c) { ___ 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: r303083 - in vendor/illumos/dist: cmd/zdb man/man1m
Author: avg Date: Wed Jul 20 09:58:10 2016 New Revision: 303083 URL: https://svnweb.freebsd.org/changeset/base/303083 Log: 7164 zdb should be able to open the root dataset illumos/illumos-gate@b702644a6eb66615d67b492fd73ecd9efa11fc7d https://github.com/illumos/illumos-gate/commit/b702644a6eb66615d67b492fd73ecd9efa11fc7d https://www.illumos.org/issues/7164 If the pool/dataset command-line argument is specified with a trailing slash, for example, "tank/", we should interpret it as the topmost dataset (rather than the whole pool) Reviewed by: Pavel Zakharov Reviewed by: Matthew Ahrens Approved by: Robert Mustacchi Author: Tim Chase Modified: vendor/illumos/dist/cmd/zdb/zdb.c vendor/illumos/dist/man/man1m/zdb.1m Modified: vendor/illumos/dist/cmd/zdb/zdb.c == --- vendor/illumos/dist/cmd/zdb/zdb.c Wed Jul 20 09:57:16 2016 (r303082) +++ vendor/illumos/dist/cmd/zdb/zdb.c Wed Jul 20 09:58:10 2016 (r303083) @@ -3530,6 +3530,7 @@ main(int argc, char **argv) uint64_t max_txg = UINT64_MAX; int rewind = ZPOOL_NEVER_REWIND; char *spa_config_path_env; + boolean_t target_is_spa = B_TRUE; (void) setrlimit(RLIMIT_NOFILE, &rl); (void) enable_extended_FILE_stdio(-1, -1); @@ -3708,8 +3709,23 @@ main(int argc, char **argv) } } + if (strpbrk(target, "/@") != NULL) { + size_t targetlen; + + target_is_spa = B_FALSE; + /* +* Remove any trailing slash. Later code would get confused +* by it, but we want to allow it so that "pool/" can +* indicate that we want to dump the topmost filesystem, +* rather than the whole pool. +*/ + targetlen = strlen(target); + if (targetlen != 0 && target[targetlen - 1] == '/') + target[targetlen - 1] = '\0'; + } + if (error == 0) { - if (strpbrk(target, "/@") == NULL || dump_opt['R']) { + if (target_is_spa || dump_opt['R']) { error = spa_open_rewind(target, &spa, FTAG, policy, NULL); if (error) { Modified: vendor/illumos/dist/man/man1m/zdb.1m == --- vendor/illumos/dist/man/man1m/zdb.1mWed Jul 20 09:57:16 2016 (r303082) +++ vendor/illumos/dist/man/man1m/zdb.1mWed Jul 20 09:58:10 2016 (r303083) @@ -56,6 +56,11 @@ pool, and is inherently unstable. The pr not documented, a knowledge of ZFS internals is assumed. .P +If the \fIdataset\fR argument does not contain any \fB/\fR or \fB@\fR +characters, it is interpreted as a pool name. The root dataset can be +specified as \fIpool\fB/\fR (pool name followed by a slash). + +.P When operating on an imported and active pool it is possible, though unlikely, that zdb may interpret inconsistent pool data and behave erratically. ___ 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: r303084 - head/cddl/contrib/opensolaris/cmd/zdb
Author: avg Date: Wed Jul 20 10:06:56 2016 New Revision: 303084 URL: https://svnweb.freebsd.org/changeset/base/303084 Log: MFV r303082: 6391 Override default SPA config location via environment illumos/illumos-gate@ae24175b2b25e9fb4bfd9ac0648b57e0735b6651 https://github.com/illumos/illumos-gate/commit/ae24175b2b25e9fb4bfd9ac0648b57e0735b6651 https://www.illumos.org/issues/6391 When using zdb with non-default SPA config file it is not convenient to add -U all the time. This commit introduces support for setting/overriding SPA config location via environment variable 'SPA_CONFIG_PATH'. If -U flag is specified in the command line it will override any other value as usual. https://github.com/zfsonlinux/zfs/commit/64d7b6cf75e52a4698d9bdec61745573c39d2e5a Reviewed by: Brian Behlendorf Reviewed by: Richard Yao Reviewed by: Matthew Ahrens Reviewed by: Will Andrews Reviewed by: George Wilson Approved by: Robert Mustacchi Author: Cyril Plisko MFC after:1 week Modified: head/cddl/contrib/opensolaris/cmd/zdb/zdb.c Directory Properties: head/cddl/contrib/opensolaris/cmd/zdb/ (props changed) Modified: head/cddl/contrib/opensolaris/cmd/zdb/zdb.c == --- head/cddl/contrib/opensolaris/cmd/zdb/zdb.c Wed Jul 20 09:58:10 2016 (r303083) +++ head/cddl/contrib/opensolaris/cmd/zdb/zdb.c Wed Jul 20 10:06:56 2016 (r303084) @@ -3558,12 +3558,22 @@ main(int argc, char **argv) nvlist_t *policy = NULL; uint64_t max_txg = UINT64_MAX; int rewind = ZPOOL_NEVER_REWIND; + char *spa_config_path_env; (void) setrlimit(RLIMIT_NOFILE, &rl); (void) enable_extended_FILE_stdio(-1, -1); dprintf_setup(&argc, argv); + /* +* If there is an environment variable SPA_CONFIG_PATH it overrides +* default spa_config_path setting. If -U flag is specified it will +* override this environment variable settings once again. +*/ + spa_config_path_env = getenv("SPA_CONFIG_PATH"); + if (spa_config_path_env != NULL) + spa_config_path = spa_config_path_env; + while ((c = getopt(argc, argv, "bcdhilmMI:suCDRSAFLXx:evp:t:U:P")) != -1) { switch (c) { ___ 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: r303085 - head/share/man/man4
Author: mav Date: Wed Jul 20 11:04:49 2016 New Revision: 303085 URL: https://svnweb.freebsd.org/changeset/base/303085 Log: Remove outdated list of supported chipsets, generalizing it. Modified: head/share/man/man4/ichsmb.4 Modified: head/share/man/man4/ichsmb.4 == --- head/share/man/man4/ichsmb.4Wed Jul 20 10:06:56 2016 (r303084) +++ head/share/man/man4/ichsmb.4Wed Jul 20 11:04:49 2016 (r303085) @@ -34,7 +34,7 @@ .\" .\" $FreeBSD$ .\" -.Dd November 28, 2007 +.Dd July 20, 2016 .Dt ICHSMB 4 .Os .Sh NAME @@ -46,10 +46,12 @@ .Cd device smb .Cd device ichsmb .Sh DESCRIPTION -This driver provides access to the SMBus controller logical -device contained in the Intel 82801AA (ICH), 82801AB (ICH0), -82801BA (ICH2), 82801CA (ICH3), 82801DC (ICH4), 82801EB (ICH5), -82801FB (ICH6) and 82801GB (ICH7) PCI chips. +The +.Nm +driver provides +.Xr smbus 4 +support for the SMBus controller logical device contained in all Intel +motherboard chipsets starting from 82801AA (ICH). .Sh SEE ALSO .Xr smb 4 , .Xr smbus 4 ___ 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: r303086 - head/cddl/contrib/opensolaris/cmd/zdb
Author: avg Date: Wed Jul 20 11:15:32 2016 New Revision: 303086 URL: https://svnweb.freebsd.org/changeset/base/303086 Log: MFV r303083: 7164 zdb should be able to open the root dataset Note: conversion of the manual page change from roff to mdoc is mine. illumos/illumos-gate@b702644a6eb66615d67b492fd73ecd9efa11fc7d https://github.com/illumos/illumos-gate/commit/b702644a6eb66615d67b492fd73ecd9efa11fc7d https://www.illumos.org/issues/7164 If the pool/dataset command-line argument is specified with a trailing slash, for example, "tank/", we should interpret it as the topmost dataset (rather than the whole pool) Reviewed by: Pavel Zakharov Reviewed by: Matthew Ahrens Approved by: Robert Mustacchi Author: Tim Chase PR: 204661 MFC after:1 week Relnotes: yes Modified: head/cddl/contrib/opensolaris/cmd/zdb/zdb.8 head/cddl/contrib/opensolaris/cmd/zdb/zdb.c Directory Properties: head/cddl/contrib/opensolaris/cmd/zdb/ (props changed) Modified: head/cddl/contrib/opensolaris/cmd/zdb/zdb.8 == --- head/cddl/contrib/opensolaris/cmd/zdb/zdb.8 Wed Jul 20 11:04:49 2016 (r303085) +++ head/cddl/contrib/opensolaris/cmd/zdb/zdb.8 Wed Jul 20 11:15:32 2016 (r303086) @@ -86,6 +86,17 @@ pool, and is inherently unstable. The precise output of most invocations is not documented, a knowledge of ZFS internals is assumed. .Pp +If the +.Ar dataset +argument does not contain any +.Sy / +or +.Sy @ +characters, it is interpreted as a pool name. +The root dataset can be specified as +.Pa pool Ns Sy / +(pool name followed by a slash). +.Pp When operating on an imported and active pool it is possible, though unlikely, that zdb may interpret inconsistent pool data and behave erratically. .Sh OPTIONS Modified: head/cddl/contrib/opensolaris/cmd/zdb/zdb.c == --- head/cddl/contrib/opensolaris/cmd/zdb/zdb.c Wed Jul 20 11:04:49 2016 (r303085) +++ head/cddl/contrib/opensolaris/cmd/zdb/zdb.c Wed Jul 20 11:15:32 2016 (r303086) @@ -3559,6 +3559,7 @@ main(int argc, char **argv) uint64_t max_txg = UINT64_MAX; int rewind = ZPOOL_NEVER_REWIND; char *spa_config_path_env; + boolean_t target_is_spa = B_TRUE; (void) setrlimit(RLIMIT_NOFILE, &rl); (void) enable_extended_FILE_stdio(-1, -1); @@ -3738,8 +3739,23 @@ main(int argc, char **argv) } } + if (strpbrk(target, "/@") != NULL) { + size_t targetlen; + + target_is_spa = B_FALSE; + /* +* Remove any trailing slash. Later code would get confused +* by it, but we want to allow it so that "pool/" can +* indicate that we want to dump the topmost filesystem, +* rather than the whole pool. +*/ + targetlen = strlen(target); + if (targetlen != 0 && target[targetlen - 1] == '/') + target[targetlen - 1] = '\0'; + } + if (error == 0) { - if (strpbrk(target, "/@") == NULL || dump_opt['R']) { + if (target_is_spa || dump_opt['R']) { error = spa_open_rewind(target, &spa, FTAG, policy, NULL); if (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"
svn commit: r303087 - head/sys/arm/allwinner
Author: manu Date: Wed Jul 20 11:23:06 2016 New Revision: 303087 URL: https://svnweb.freebsd.org/changeset/base/303087 Log: Add support for the SID (Security ID Module) on Allwinner A10 and A20. The rootkey is burnt at production and can't be changed, thus is can be used as a device unique ID or to generate a MAC address (This is was u-boot does). The rootkey is exposed as a sysctl (dev.aw_sid..rootkey). Reviewed by: jmcneill Differential Revision:https://reviews.freebsd.org/D6383 Modified: head/sys/arm/allwinner/aw_sid.c head/sys/arm/allwinner/aw_sid.h Modified: head/sys/arm/allwinner/aw_sid.c == --- head/sys/arm/allwinner/aw_sid.c Wed Jul 20 11:15:32 2016 (r303086) +++ head/sys/arm/allwinner/aw_sid.c Wed Jul 20 11:23:06 2016 (r303087) @@ -33,12 +33,14 @@ #include __FBSDID("$FreeBSD$"); +#include #include #include #include #include #include #include +#include #include #include @@ -50,13 +52,22 @@ __FBSDID("$FreeBSD$"); #defineSID_THERMAL_CALIB0 (SID_SRAM + 0x34) #defineSID_THERMAL_CALIB1 (SID_SRAM + 0x38) +enum sid_type { + A10_SID = 1, + A20_SID, + A83T_SID, +}; + static struct ofw_compat_data compat_data[] = { - { "allwinner,sun8i-a83t-sid", 1 }, + { "allwinner,sun4i-a10-sid",A10_SID}, + { "allwinner,sun7i-a20-sid",A20_SID}, + { "allwinner,sun8i-a83t-sid", A83T_SID}, { NULL, 0 } }; struct aw_sid_softc { struct resource *res; + int type; }; static struct aw_sid_softc *aw_sid_sc; @@ -66,9 +77,18 @@ static struct resource_spec aw_sid_spec[ { -1, 0 } }; +enum sid_keys { + AW_SID_ROOT_KEY, +}; + +#defineROOT_KEY_OFF0x0 +#defineROOT_KEY_SIZE 4 + #defineRD4(sc, reg)bus_read_4((sc)->res, (reg)) #defineWR4(sc, reg, val) bus_write_4((sc)->res, (reg), (val)) +static int aw_sid_sysctl(SYSCTL_HANDLER_ARGS); + static int aw_sid_probe(device_t dev) { @@ -96,6 +116,19 @@ aw_sid_attach(device_t dev) aw_sid_sc = sc; + sc->type = ofw_bus_search_compatible(dev, compat_data)->ocd_data; + switch (sc->type) { + case A10_SID: + case A20_SID: + SYSCTL_ADD_PROC(device_get_sysctl_ctx(dev), + SYSCTL_CHILDREN(device_get_sysctl_tree(dev)), + OID_AUTO, "rootkey", + CTLTYPE_STRING | CTLFLAG_RD, + dev, AW_SID_ROOT_KEY, aw_sid_sysctl, "A", "Root Key"); + break; + default: + break; + } return (0); } @@ -107,6 +140,8 @@ aw_sid_read_tscalib(uint32_t *calib0, ui sc = aw_sid_sc; if (sc == NULL) return (ENXIO); + if (sc->type != A83T_SID) + return (ENXIO); *calib0 = RD4(sc, SID_THERMAL_CALIB0); *calib1 = RD4(sc, SID_THERMAL_CALIB1); @@ -114,6 +149,45 @@ aw_sid_read_tscalib(uint32_t *calib0, ui return (0); } +int +aw_sid_get_rootkey(u_char *out) +{ + struct aw_sid_softc *sc; + int i; + u_int tmp; + + sc = aw_sid_sc; + if (sc == NULL) + return (ENXIO); + if (sc->type != A10_SID && sc->type != A20_SID) + return (ENXIO); + + for (i = 0; i < ROOT_KEY_SIZE ; i++) { + tmp = RD4(aw_sid_sc, ROOT_KEY_OFF + (i * 4)); + be32enc(&out[i * 4], tmp); + } + + return (0); +} + +static int +aw_sid_sysctl(SYSCTL_HANDLER_ARGS) +{ + enum sid_keys key = arg2; + u_char rootkey[16]; + char out[33]; + + if (key != AW_SID_ROOT_KEY) + return (ENOENT); + + if (aw_sid_get_rootkey(rootkey) != 0) + return (ENOENT); + snprintf(out, sizeof(out), + "%16D", rootkey, ""); + + return sysctl_handle_string(oidp, out, sizeof(out), req); +} + static device_method_t aw_sid_methods[] = { /* Device interface */ DEVMETHOD(device_probe, aw_sid_probe), Modified: head/sys/arm/allwinner/aw_sid.h == --- head/sys/arm/allwinner/aw_sid.h Wed Jul 20 11:15:32 2016 (r303086) +++ head/sys/arm/allwinner/aw_sid.h Wed Jul 20 11:23:06 2016 (r303087) @@ -30,5 +30,6 @@ #define __AW_SID_H__ intaw_sid_read_tscalib(uint32_t *, uint32_t *); +intaw_sid_get_rootkey(u_char *out); #endif /* !__AW_SID_H__ */ ___ 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: r301453 - in head/sys: arm/arm arm64/arm64 dev/fdt dev/gpio dev/iicbus dev/ofw dev/pci dev/vnic kern mips/mips sys
Dne 19.07.2016 v 17:06 Nathan Whitehorn napsal(a): > > > On 07/19/16 04:13, Michal Meloun wrote: >> Dne 19.07.2016 v 2:11 Nathan Whitehorn napsal(a): >> Hi Nathan, >> I’m afraid that skra is on vacation, for next 2 weeks (at minimum), so >> please don’t expect quick response. >> >>> Could you please describe what this change is in more detail? >> Short description is appended. >> >>> It breaks a lot of encapsulations we have worked very hard to maintain, >>> moves ARM code into MI parts of the kernel, and the OFW parts violate >>> IEEE 1275 (the Open Firmware standard). In particular, there is no >>> guarantee that the interrupts for a newbus (or OF) device are >>> encoded in >>> a property called "interrupts" (or, indeed, in any property at all) on >>> that node and there are many, many device trees where that is not the >>> case (e.g. ones with interrupt maps, as well as Apple hardware). By >>> putting that knowledge into the OF root bus device, which we have tried >>> to keep it out of, this enforces a standard that doesn't actually >>> exist. >> Imho, this patch doesn’t change anything in this area. Only handling of >> “interrupts” property is changed, all other cases are unchanged (I >> hope). Also, INTRNG code is currently shared by ARM, ARM64 and MIPS. > > But "interrupts" isn't a generic part of OF. This makes it one, > incorrectly. How? Can you be little more exact ? > >> >>> I'm hesitant to ask for reversion on something that landed 6 weeks ago >>> without me noticing, but this needs a lot more architectural work >>> before >>> any parts of the kernel should use it. >>> -Nathan >> I think that it’s too late. This patch series consist of r301451 >> (https://reviews.freebsd.org/D6632), >> r301453, r301539 and 301543. And new GPIO interrupts are currently used >> (by in tree drivers or in development trees). > > Well, then we need in-place rearchitecture. > >> >> >> The root of problem is that standard way of delivering interrupt >> resource to consumer driver doesn’t works in OFW world. >> >> So we have some fact: >> - the format of interrupt property is dependent of interrupt >>controller and only interrupt controller can parse it. >> - the interrupt property can have more data than just interrupt >>number. >> - single interrupt controller must be able to handle multiple >>format of interrupt description. >> >> In pre-patchset era, simplebus enumerates children and attempts to set >> memory and interrupts to resource list for them. But the interrupt >> controllers are not yet populated so nobody can parse interrupt >> property. Moreover, in all cases (parsed or not), we cannot store >> complete interrupt description into resource list. > > We have done this for many years on PowerPC and sparc64 with delayed > configuration of interrupts and a look-up table. This handles > complicated bus configurations better than this code and requires no > changes outside of a few MD files. That is why the (now partially > duplicated) OFW_BUS_MAP_INTR() function exists. That one also has the > benefit of still working when used in conjunction with, e.g., devices > with an interrupt-map-mask property. > >> >> The patch simply postpones reading of interrupt property to >> bus_alloc_resource() (called by consumer driver) time. >> >> Due to this, we can: >> - parse interrupt property. The interrupt driver must exist >>at this time. > > This only works with some types of interrupt properties, not all, and > breaks if the interrupt driver hasn't attached yet (which it can't be > guaranteed to -- some PPC systems have interrupt drivers that live on > the PCI bus, for example). How you can allocate (and reserve it in rman) interrupt if is not mapped (so you have not real IRQ number for it). Just for notice - multiple virtual IRQs can be mapped into single real IRQ. > >> - bus_alloc_resource() returns resource, so we can attach parsed >>interrupt data to it. By this, the resource itself can be used >>for delivering configuration data to subsequent call to >>bus_setup_intr() (or to all related bus_() calls). >> >> >> The patched code still accepts delivering of interrupts in resource >> list. >> >> Michal >> > > Given that other code depends on this, fixing it will likely require > some complex work. I wish I had known about it when it went in. > > There are three main problems: > 1. It doesn't work for interrupts defined by other mechanisms (e.g. > interrupt-map properties) I aggree, but missing ' interrupt-map' functioanlity is not caused by this patch. > 2. It partially duplicates the functionality of OFW_BUS_MAP_INTR(), > but is both problematically more general and less flexible (it has > requirements on timing of PIC attachment vs. driver resource allocation) OFW_BUS_MAP_INTR() can parse only OFW based data and expect that parsed data are magicaly stored within the call. The new method, bus_map_intr(), can parse data from multiple sources (OFW, UEFI / ACPI, synthetic[gpio dev
svn commit: r303088 - head/lib/libc/gen
Author: ache Date: Wed Jul 20 12:46:21 2016 New Revision: 303088 URL: https://svnweb.freebsd.org/changeset/base/303088 Log: In addition to r303074 case 1, search for protected L'/' too in globtilde() Modified: head/lib/libc/gen/glob.c Modified: head/lib/libc/gen/glob.c == --- head/lib/libc/gen/glob.cWed Jul 20 11:23:06 2016(r303087) +++ head/lib/libc/gen/glob.cWed Jul 20 12:46:21 2016(r303088) @@ -126,7 +126,6 @@ struct glob_limit { #defineTILDE L'~' #defineLBRACE L'{' #defineRBRACE L'}' -#defineSLASH L'/' #defineCOMMA L',' #ifndef DEBUG @@ -427,10 +426,10 @@ globtilde(const Char *pattern, Char *pat */ eb = &patbuf[patbuf_len - 1]; for (p = pattern + 1, b = patbuf; - b < eb && *p != EOS && *p != SLASH; *b++ = *p++) + b < eb && *p != EOS && UNPROT(*p) != SEP; *b++ = *p++) continue; - if (*p != EOS && *p != SLASH) + if (*p != EOS && UNPROT(*p) != SEP) return (NULL); *b = EOS; ___ 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: r303037 - head/sys/kern
Gleb I wish you would have responded earlier.. I am more than glad to hand off all kern_timeout.c to you… please take it commit what you want to it and have it. I hate the code and I dislike having to touch it. Its yours.. I can assure you I will not touch it again. R > On Jul 20, 2016, at 8:53 AM, Gleb Smirnoff wrote: > > On Tue, Jul 19, 2016 at 06:31:19PM +, Randall Stewart wrote: > R> Author: rrs > R> Date: Tue Jul 19 18:31:19 2016 > R> New Revision: 303037 > R> URL: https://svnweb.freebsd.org/changeset/base/303037 > R> > R> Log: > R> This reverts out Gleb's changes and adds three small > R> fixes that I think closes up the races Gleb was > R> looking for. This is running quite nicely in Netflix and > R> now no longer causes TCP-tcb leaks. > R> > R> Differential Revision: 7135 > > Just to notice that I am completely pissed of by this commit > war, that you started. > > I've been testing my changes properly, I gave people time to > review my changes. You didn't. > > From your explanation in other emails I see that you've been > testing your changes with a version of FreeBSD that is a heavily > modified FreeBSD 10, not 11. > > The new code you mixed with revert of mine, doesn't fix the > problem observed. It fixes another problem that you imagined, > which might exist, but isn't observed. We already discussed that > and you didn't prove it wrong. > > Your change doesn't even revert my change completely. > > -- > Totus tuus, Glebius. Randall Stewart r...@netflix.com 803-317-4952 ___ 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: r303027 - in stable/11/release: . scripts
Hi, I made this PR for the issue. https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=211029 Ronald. On Tue, 19 Jul 2016 18:22:50 +0200, Glen Barber wrote: Author: gjb Date: Tue Jul 19 16:22:50 2016 New Revision: 303027 URL: https://svnweb.freebsd.org/changeset/base/303027 Log: Reduce the disc1.iso size from 850+M to just over 650M. As a result of this change, the 'kernel-dbg.txz' distribution is no longer provided on disc1.iso, and deselected by default in bsdinstall(8). When 'kernel-dbg.txz' is selected, network configuration happens before the installer proceeds, to fetch the distribution from the mirrors. This is a direct commit to stable/11, as there is intention to solve this differently for 12.0-RELEASE. Reviewed by: nwhitehorn (glanced at) Approved by: re (hrs) Sponsored by: The FreeBSD Foundation Modified: stable/11/release/Makefile stable/11/release/scripts/make-manifest.sh Modified: stable/11/release/Makefile == --- stable/11/release/Makefile Tue Jul 19 16:02:07 2016(r303026) +++ stable/11/release/Makefile Tue Jul 19 16:22:50 2016(r303027) @@ -172,11 +172,12 @@ disc1: packagesystem mkdir -p ${.TARGET} cd ${WORLDDIR} && ${IMAKE} installkernel installworld distribution \ DESTDIR=${.OBJDIR}/${.TARGET} MK_RESCUE=no MK_KERNEL_SYMBOLS=no \ - MK_PROFILE=no MK_SENDMAIL=no MK_TESTS=no MK_LIB32=no \ - MK_DEBUG_FILES=no + MK_PROFILE=no MK_MAIL=no MK_TESTS=no MK_LIB32=no \ + MK_DEBUG_FILES=no MK_LLDB=no \ + MK_TOOLCHAIN=no # Copy distfiles mkdir -p ${.TARGET}/usr/freebsd-dist - for dist in MANIFEST $$(ls *.txz | grep -vE -- '(base|lib32)-dbg'); \ + for dist in MANIFEST $$(ls *.txz | grep -vE -- '(base|lib32|kernel)-dbg'); \ do cp $${dist} ${.TARGET}/usr/freebsd-dist; \ done # Copy documentation, if generated Modified: stable/11/release/scripts/make-manifest.sh == --- stable/11/release/scripts/make-manifest.sh Tue Jul 19 16:02:07 2016 (r303026) +++ stable/11/release/scripts/make-manifest.sh Tue Jul 19 16:22:50 2016 (r303027) @@ -36,7 +36,7 @@ default_tests=off default_base_dbg=off default_lib32_dbg=off default_kernel_alt=off -default_kernel_dbg=on +default_kernel_dbg=off default_kernel_alt_dbg=off for i in ${*}; do ___ 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-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: r303027 - in stable/11/release: . scripts
My apologies for not mentioning that in the commit log. I'd like to keep the PR open, though, since we're going to fix this differently for current. Glen On Wed, Jul 20, 2016 at 03:46:24PM +0200, Ronald Klop wrote: > Hi, I made this PR for the issue. > > https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=211029 > > Ronald. > > > > On Tue, 19 Jul 2016 18:22:50 +0200, Glen Barber wrote: > > >Author: gjb > >Date: Tue Jul 19 16:22:50 2016 > >New Revision: 303027 > >URL: https://svnweb.freebsd.org/changeset/base/303027 > > > >Log: > > Reduce the disc1.iso size from 850+M to just over 650M. > > As a result of this change, the 'kernel-dbg.txz' distribution > > is no longer provided on disc1.iso, and deselected by default > > in bsdinstall(8). When 'kernel-dbg.txz' is selected, network > > configuration happens before the installer proceeds, to fetch > > the distribution from the mirrors. > > This is a direct commit to stable/11, as there is intention > > to solve this differently for 12.0-RELEASE. > > Reviewed by:nwhitehorn (glanced at) > > Approved by: re (hrs) > > Sponsored by: The FreeBSD Foundation > > > >Modified: > > stable/11/release/Makefile > > stable/11/release/scripts/make-manifest.sh > > > >Modified: stable/11/release/Makefile > >== > >--- stable/11/release/Makefile Tue Jul 19 16:02:07 2016 > >(r303026) > >+++ stable/11/release/Makefile Tue Jul 19 16:22:50 2016 > >(r303027) > >@@ -172,11 +172,12 @@ disc1: packagesystem > > mkdir -p ${.TARGET} > > cd ${WORLDDIR} && ${IMAKE} installkernel installworld distribution \ > > DESTDIR=${.OBJDIR}/${.TARGET} MK_RESCUE=no MK_KERNEL_SYMBOLS=no > > \ > >-MK_PROFILE=no MK_SENDMAIL=no MK_TESTS=no MK_LIB32=no \ > >-MK_DEBUG_FILES=no > >+MK_PROFILE=no MK_MAIL=no MK_TESTS=no MK_LIB32=no \ > >+MK_DEBUG_FILES=no MK_LLDB=no \ > >+MK_TOOLCHAIN=no > > # Copy distfiles > > mkdir -p ${.TARGET}/usr/freebsd-dist > >-for dist in MANIFEST $$(ls *.txz | grep -vE -- '(base|lib32)-dbg'); \ > >+for dist in MANIFEST $$(ls *.txz | grep -vE -- > >'(base|lib32|kernel)-dbg'); \ > > do cp $${dist} ${.TARGET}/usr/freebsd-dist; \ > > done > > # Copy documentation, if generated > > > >Modified: stable/11/release/scripts/make-manifest.sh > >== > >--- stable/11/release/scripts/make-manifest.sh Tue Jul 19 16:02:07 > >2016 (r303026) > >+++ stable/11/release/scripts/make-manifest.sh Tue Jul 19 16:22:50 > >2016 (r303027) > >@@ -36,7 +36,7 @@ default_tests=off > > default_base_dbg=off > > default_lib32_dbg=off > > default_kernel_alt=off > >-default_kernel_dbg=on > >+default_kernel_dbg=off > > default_kernel_alt_dbg=off > >for i in ${*}; do > >___ > >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" > signature.asc Description: PGP signature
svn commit: r303089 - in stable/10: share/man/man4 sys/dev/mpr sys/dev/mps
Author: slm Date: Wed Jul 20 14:38:02 2016 New Revision: 303089 URL: https://svnweb.freebsd.org/changeset/base/303089 Log: MFC r302673 Use real values to calculate Max I/O size instead of guessing. Reviewed by: ken, scottl Approved by: ken, scottl, ambrisko (mentors) Differential Revision:https://reviews.freebsd.org/D7043 Modified: stable/10/share/man/man4/mpr.4 stable/10/share/man/man4/mps.4 stable/10/sys/dev/mpr/mpr.c stable/10/sys/dev/mpr/mpr_sas.c stable/10/sys/dev/mpr/mprvar.h stable/10/sys/dev/mps/mps.c stable/10/sys/dev/mps/mps_sas.c stable/10/sys/dev/mps/mpsvar.h Directory Properties: stable/10/ (props changed) Modified: stable/10/share/man/man4/mpr.4 == --- stable/10/share/man/man4/mpr.4 Wed Jul 20 12:46:21 2016 (r303088) +++ stable/10/share/man/man4/mpr.4 Wed Jul 20 14:38:02 2016 (r303089) @@ -29,7 +29,7 @@ .\" 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 DAMAGES. -.\" +.\" .\" mpr driver man page. .\" .\" Author: Ken Merry @@ -38,7 +38,7 @@ .\" $Id$ .\" $FreeBSD$ .\" -.Dd April 29, 2016 +.Dd July 6, 2016 .Dt MPR 4 .Os .Sh NAME @@ -156,6 +156,29 @@ The current number of active I/O command dev.mpr.X.io_cmds_active .Xr sysctl 8 variable. +.Ed +.Pp +To set the maximum number of pages that will be used per I/O for all adapters, +set this tunable in +.Xr loader.conf 5 : +.Bd -literal -offset indent +hw.mpr.max_io_pages= +.Ed +.Pp +To set the maximum number of pages that will be used per I/O for a specific +adapter, set this tunable in +.Xr loader.conf 5 : +.Bd -literal -offset indent +dev.mpr.X.max_io_pages= +.Ed +.Pp +The default max_io_pages value is -1, meaning that the maximum I/O size that +will be used per I/O will be calculated using the IOCFacts values stored in +the controller. +The lowest value that the driver will use for max_io_pages is 1, otherwise +IOCFacts will be used to calculate the maximum I/O size. +The smaller I/O size calculated from either max_io_pages or IOCFacts will be the +maximum I/O size used by the driver. .Pp The highest number of active I/O commands seen since boot is stored in the dev.mpr.X.io_cmds_highwater @@ -220,7 +243,7 @@ SATA disks that take several seconds to command might not be discovered by the driver. This problem can sometimes be overcome by increasing the value of the spinup wait time in -.Xr loader.conf 5 : +.Xr loader.conf 5 with the .Bd -literal -offset indent hw.mpr.spinup_wait_time= @@ -250,7 +273,7 @@ hw.mpr.X.debug_level .Pp tunable, either in .Xr loader.conf 5 -or by using +or by using .Xr sysctl 8 . These bits have the described effects: .Bd -literal -offset indent Modified: stable/10/share/man/man4/mps.4 == --- stable/10/share/man/man4/mps.4 Wed Jul 20 12:46:21 2016 (r303088) +++ stable/10/share/man/man4/mps.4 Wed Jul 20 14:38:02 2016 (r303089) @@ -1,5 +1,8 @@ .\" .\" Copyright (c) 2010 Spectra Logic Corporation +.\" Copyright (c) 2014 LSI Corp +.\" Copyright (c) 2016 Avago Technologies +.\" Copyright (c) 2016 Broadcom Ltd. .\" All rights reserved. .\" .\" Redistribution and use in source and binary forms, with or without @@ -30,25 +33,27 @@ .\" mps driver man page. .\" .\" Author: Ken Merry +.\" Author: Stephen McConnell .\" .\" $Id: //depot/SpectraBSD/head/share/man/man4/mps.4#6 $ .\" $FreeBSD$ .\" -.Dd December 9, 2015 +.Dd July 5, 2016 .Dt MPS 4 .Os .Sh NAME .Nm mps -.Nd LSI Fusion-MPT 2 Serial Attached SCSI driver +.Nd "LSI Fusion-MPT 2 IT/IR 6Gb/s Serial Attached SCSI/SATA driver" .Sh SYNOPSIS -To compile this driver into your kernel, -place the following lines in your kernel configuration file: +To compile this driver into the kernel, place these lines in the kernel +configuration file: .Bd -ragged -offset indent +.Cd "device pci" .Cd "device scbus" .Cd "device mps" .Ed .Pp -Or, to load the driver as a module at boot, place the following line in +The driver can be loaded as a module at boot time by placing this line in .Xr loader.conf 5 : .Bd -literal -offset indent mps_load="YES" @@ -56,35 +61,30 @@ mps_load="YES" .Sh DESCRIPTION The .Nm -driver provides support for LSI Logic Fusion-MPT 2 +driver provides support for Broadcom Ltd./Avago Tech (LSI) +Fusion-MPT 2 IT/IR .Tn SAS controllers and WarpDrive solid state storage cards. .Sh HARDWARE -The +These controllers are supported by the .Nm -driver supports the following hardware: +driver: .Pp .Bl -bullet -compact .It -LSI Logic SAS2004 (4 Port -.Tn SAS ) +Broadcom Ltd./Avago Tech (LSI) SAS 2004 (4 Port SAS) .It -LSI Logic SAS2008 (8 Port -.Tn SAS ) +Broadcom Ltd./Avago Tech (LSI) SAS 2008 (8 Port SAS) .It -LSI Logic SAS2108 (8 Port -.Tn SAS
svn commit: r303090 - head/sys/ufs/ufs
Author: kib Date: Wed Jul 20 14:40:56 2016 New Revision: 303090 URL: https://svnweb.freebsd.org/changeset/base/303090 Log: Ensure that the UFS directory vnode' vm_object is properly sized before UFS_BALLOC() is called. I do not believe that this caused any real issue on FreeBSD because the exclusive vnode lock is held over the balloc/resize, the change is to make formally correct KPI use. Based on: the Matthew Dillon' patch from DragonFly BSD PR: 93942 Reviewed by: mckusick Tested by:pho Sponsored by: The FreeBSD Foundation MFC after:1 week Modified: head/sys/ufs/ufs/ufs_lookup.c head/sys/ufs/ufs/ufs_vnops.c Modified: head/sys/ufs/ufs/ufs_lookup.c == --- head/sys/ufs/ufs/ufs_lookup.c Wed Jul 20 14:38:02 2016 (r303089) +++ head/sys/ufs/ufs/ufs_lookup.c Wed Jul 20 14:40:56 2016 (r303090) @@ -881,6 +881,7 @@ ufs_direnter(dvp, tvp, dirp, cnp, newdir struct buf *bp; u_int dsize; struct direct *ep, *nep; + u_int64_t old_isize; int error, ret, blkoff, loc, spacefree, flags, namlen; char *dirbuf; @@ -909,16 +910,18 @@ ufs_direnter(dvp, tvp, dirp, cnp, newdir return (error); } #endif + old_isize = dp->i_size; + vnode_pager_setsize(dvp, (u_long)dp->i_offset + DIRBLKSIZ); if ((error = UFS_BALLOC(dvp, (off_t)dp->i_offset, DIRBLKSIZ, cr, flags, &bp)) != 0) { if (DOINGSOFTDEP(dvp) && newdirbp != NULL) bdwrite(newdirbp); + vnode_pager_setsize(dvp, (u_long)old_isize); return (error); } dp->i_size = dp->i_offset + DIRBLKSIZ; DIP_SET(dp, i_size, dp->i_size); dp->i_flag |= IN_CHANGE | IN_UPDATE; - vnode_pager_setsize(dvp, (u_long)dp->i_size); dirp->d_reclen = DIRBLKSIZ; blkoff = dp->i_offset & (VFSTOUFS(dvp->v_mount)->um_mountp->mnt_stat.f_iosize - 1); Modified: head/sys/ufs/ufs/ufs_vnops.c == --- head/sys/ufs/ufs/ufs_vnops.cWed Jul 20 14:38:02 2016 (r303089) +++ head/sys/ufs/ufs/ufs_vnops.cWed Jul 20 14:40:56 2016 (r303090) @@ -1913,13 +1913,13 @@ ufs_mkdir(ap) dirtemplate = *dtp; dirtemplate.dot_ino = ip->i_number; dirtemplate.dotdot_ino = dp->i_number; + vnode_pager_setsize(tvp, DIRBLKSIZ); if ((error = UFS_BALLOC(tvp, (off_t)0, DIRBLKSIZ, cnp->cn_cred, BA_CLRBUF, &bp)) != 0) goto bad; ip->i_size = DIRBLKSIZ; DIP_SET(ip, i_size, DIRBLKSIZ); ip->i_flag |= IN_CHANGE | IN_UPDATE; - vnode_pager_setsize(tvp, (u_long)ip->i_size); bcopy((caddr_t)&dirtemplate, (caddr_t)bp->b_data, sizeof dirtemplate); if (DOINGSOFTDEP(tvp)) { /* ___ 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: r303091 - head/usr.sbin/camdd
Author: asomers Date: Wed Jul 20 15:00:05 2016 New Revision: 303091 URL: https://svnweb.freebsd.org/changeset/base/303091 Log: Fix camdd for host-aware and host-managed SMR disks. This should've gone in as part of r300207 Reviewed by: ken MFC after:3 days Sponsored by: Spectra Logic Corp Modified: head/usr.sbin/camdd/camdd.c Modified: head/usr.sbin/camdd/camdd.c == --- head/usr.sbin/camdd/camdd.c Wed Jul 20 14:40:56 2016(r303090) +++ head/usr.sbin/camdd/camdd.c Wed Jul 20 15:00:05 2016(r303091) @@ -1292,6 +1292,7 @@ camdd_probe_pass(struct cam_device *cam_ case T_CDROM: case T_OPTICAL: case T_RBC: + case T_ZBC_HM: break; default: errx(1, "Unsupported SCSI device type %d", scsi_dev_type); ___ 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: r303092 - in stable/11: sys/kern sys/sys usr.bin/kdump
Author: kib Date: Wed Jul 20 15:02:37 2016 New Revision: 303092 URL: https://svnweb.freebsd.org/changeset/base/303092 Log: MFC r302770: Trace timeval parameters to the getitimer(2) and setitimer(2) syscalls. Approved by: re (gjb) Modified: stable/11/sys/kern/kern_time.c stable/11/sys/sys/ktrace.h stable/11/usr.bin/kdump/kdump.c Directory Properties: stable/11/ (props changed) Modified: stable/11/sys/kern/kern_time.c == --- stable/11/sys/kern/kern_time.c Wed Jul 20 15:00:05 2016 (r303091) +++ stable/11/sys/kern/kern_time.c Wed Jul 20 15:02:37 2016 (r303092) @@ -32,6 +32,8 @@ #include __FBSDID("$FreeBSD$"); +#include "opt_ktrace.h" + #include #include #include @@ -54,6 +56,9 @@ __FBSDID("$FreeBSD$"); #include #include #include +#ifdef KTRACE +#include +#endif #include #include @@ -701,6 +706,10 @@ kern_getitimer(struct thread *td, u_int *aitv = p->p_stats->p_timer[which]; PROC_ITIMUNLOCK(p); } +#ifdef KTRACE + if (KTRPOINT(td, KTR_STRUCT)) + ktritimerval(aitv); +#endif return (0); } @@ -742,6 +751,10 @@ kern_setitimer(struct thread *td, u_int if (which > ITIMER_PROF) return (EINVAL); +#ifdef KTRACE + if (KTRPOINT(td, KTR_STRUCT)) + ktritimerval(aitv); +#endif if (itimerfix(&aitv->it_value) || aitv->it_value.tv_sec > INT32_MAX / 2) return (EINVAL); @@ -786,6 +799,10 @@ kern_setitimer(struct thread *td, u_int p->p_stats->p_timer[which] = *aitv; PROC_ITIMUNLOCK(p); } +#ifdef KTRACE + if (KTRPOINT(td, KTR_STRUCT)) + ktritimerval(oitv); +#endif return (0); } Modified: stable/11/sys/sys/ktrace.h == --- stable/11/sys/sys/ktrace.h Wed Jul 20 15:00:05 2016(r303091) +++ stable/11/sys/sys/ktrace.h Wed Jul 20 15:02:37 2016(r303092) @@ -270,6 +270,8 @@ voidktrcapfail(enum ktr_cap_fail_type, const cap_rights_t *); #define ktrcaprights(s) \ ktrstruct("caprights", (s), sizeof(cap_rights_t)) +#definektritimerval(s) \ + ktrstruct("itimerval", (s), sizeof(struct itimerval)) #define ktrsockaddr(s) \ ktrstruct("sockaddr", (s), ((struct sockaddr *)(s))->sa_len) #define ktrstat(s) \ Modified: stable/11/usr.bin/kdump/kdump.c == --- stable/11/usr.bin/kdump/kdump.c Wed Jul 20 15:00:05 2016 (r303091) +++ stable/11/usr.bin/kdump/kdump.c Wed Jul 20 15:02:37 2016 (r303092) @@ -109,6 +109,7 @@ void ktruser_malloc(void *); void ktruser_rtld(int, void *); void ktruser(int, void *); void ktrcaprights(cap_rights_t *); +void ktritimerval(struct itimerval *it); void ktrsockaddr(struct sockaddr *); void ktrstat(struct stat *); void ktrstruct(char *, size_t); @@ -1562,6 +1563,24 @@ ktrcaprights(cap_rights_t *rightsp) printf("\n"); } +static void +ktrtimeval(struct timeval *tv) +{ + + printf("{%ld, %ld}", (long)tv->tv_sec, tv->tv_usec); +} + +void +ktritimerval(struct itimerval *it) +{ + + printf("itimerval { .interval = "); + ktrtimeval(&it->it_interval); + printf(", .value = "); + ktrtimeval(&it->it_value); + printf(" }\n"); +} + void ktrsockaddr(struct sockaddr *sa) { @@ -1737,6 +1756,7 @@ ktrstruct(char *buf, size_t buflen) size_t namelen, datalen; int i; cap_rights_t rights; + struct itimerval it; struct stat sb; struct sockaddr_storage ss; @@ -1761,6 +1781,11 @@ ktrstruct(char *buf, size_t buflen) goto invalid; memcpy(&rights, data, datalen); ktrcaprights(&rights); + } else if (strcmp(name, "itimerval") == 0) { + if (datalen != sizeof(struct itimerval)) + goto invalid; + memcpy(&it, data, datalen); + ktritimerval(&it); } else if (strcmp(name, "stat") == 0) { if (datalen != sizeof(struct stat)) goto invalid; ___ 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: r303093 - in stable/10: sys/kern sys/sys usr.bin/kdump
Author: kib Date: Wed Jul 20 15:07:52 2016 New Revision: 303093 URL: https://svnweb.freebsd.org/changeset/base/303093 Log: MFC r302770: Trace timeval parameters to the getitimer(2) and setitimer(2) syscalls. Modified: stable/10/sys/kern/kern_time.c stable/10/sys/sys/ktrace.h stable/10/usr.bin/kdump/kdump.c Directory Properties: stable/10/ (props changed) Modified: stable/10/sys/kern/kern_time.c == --- stable/10/sys/kern/kern_time.c Wed Jul 20 15:02:37 2016 (r303092) +++ stable/10/sys/kern/kern_time.c Wed Jul 20 15:07:52 2016 (r303093) @@ -32,6 +32,8 @@ #include __FBSDID("$FreeBSD$"); +#include "opt_ktrace.h" + #include #include #include @@ -54,6 +56,9 @@ __FBSDID("$FreeBSD$"); #include #include #include +#ifdef KTRACE +#include +#endif #include #include @@ -699,6 +704,10 @@ kern_getitimer(struct thread *td, u_int *aitv = p->p_stats->p_timer[which]; PROC_ITIMUNLOCK(p); } +#ifdef KTRACE + if (KTRPOINT(td, KTR_STRUCT)) + ktritimerval(aitv); +#endif return (0); } @@ -740,6 +749,10 @@ kern_setitimer(struct thread *td, u_int if (which > ITIMER_PROF) return (EINVAL); +#ifdef KTRACE + if (KTRPOINT(td, KTR_STRUCT)) + ktritimerval(aitv); +#endif if (itimerfix(&aitv->it_value) || aitv->it_value.tv_sec > INT32_MAX / 2) return (EINVAL); @@ -784,6 +797,10 @@ kern_setitimer(struct thread *td, u_int p->p_stats->p_timer[which] = *aitv; PROC_ITIMUNLOCK(p); } +#ifdef KTRACE + if (KTRPOINT(td, KTR_STRUCT)) + ktritimerval(oitv); +#endif return (0); } Modified: stable/10/sys/sys/ktrace.h == --- stable/10/sys/sys/ktrace.h Wed Jul 20 15:02:37 2016(r303092) +++ stable/10/sys/sys/ktrace.h Wed Jul 20 15:07:52 2016(r303093) @@ -270,6 +270,8 @@ voidktrcapfail(enum ktr_cap_fail_type, const cap_rights_t *); #define ktrcaprights(s) \ ktrstruct("caprights", (s), sizeof(cap_rights_t)) +#definektritimerval(s) \ + ktrstruct("itimerval", (s), sizeof(struct itimerval)) #define ktrsockaddr(s) \ ktrstruct("sockaddr", (s), ((struct sockaddr *)(s))->sa_len) #define ktrstat(s) \ Modified: stable/10/usr.bin/kdump/kdump.c == --- stable/10/usr.bin/kdump/kdump.c Wed Jul 20 15:02:37 2016 (r303092) +++ stable/10/usr.bin/kdump/kdump.c Wed Jul 20 15:07:52 2016 (r303093) @@ -106,6 +106,7 @@ void ktruser_malloc(void *); void ktruser_rtld(int, void *); void ktruser(int, void *); void ktrcaprights(cap_rights_t *); +void ktritimerval(struct itimerval *it); void ktrsockaddr(struct sockaddr *); void ktrstat(struct stat *); void ktrstruct(char *, size_t); @@ -1616,6 +1617,24 @@ ktrcaprights(cap_rights_t *rightsp) printf("\n"); } +static void +ktrtimeval(struct timeval *tv) +{ + + printf("{%ld, %ld}", (long)tv->tv_sec, tv->tv_usec); +} + +void +ktritimerval(struct itimerval *it) +{ + + printf("itimerval { .interval = "); + ktrtimeval(&it->it_interval); + printf(", .value = "); + ktrtimeval(&it->it_value); + printf(" }\n"); +} + void ktrsockaddr(struct sockaddr *sa) { @@ -1799,6 +1818,7 @@ ktrstruct(char *buf, size_t buflen) size_t namelen, datalen; int i; cap_rights_t rights; + struct itimerval it; struct stat sb; struct sockaddr_storage ss; @@ -1823,6 +1843,11 @@ ktrstruct(char *buf, size_t buflen) goto invalid; memcpy(&rights, data, datalen); ktrcaprights(&rights); + } else if (strcmp(name, "itimerval") == 0) { + if (datalen != sizeof(struct itimerval)) + goto invalid; + memcpy(&it, data, datalen); + ktritimerval(&it); } else if (strcmp(name, "stat") == 0) { if (datalen != sizeof(struct stat)) goto invalid; ___ 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: r303043 - in head: share/man/man4 sys/dev/vt sys/dev/vt/hw/fb
On Wed, Jul 20, 2016 at 12:44 AM, Kubilay Kocak wrote: > On 20/07/2016 5:20 AM, Conrad E. Meyer wrote: >> Author: cem >> Date: Tue Jul 19 19:20:47 2016 >> New Revision: 303043 >> URL: https://svnweb.freebsd.org/changeset/base/303043 >> >> Log: >> Increase vt(4) framebuffer maximum size >> >> And rename "DEFAULT" constants to the more accurate "MAX." >> >> PR: 210382 >> Submitted by: Felix >> Reviewed by:wblock, cem >> Tested by: Dave Cottlehuber > > Can MFC? > > There are high hopes for this coming out in 11.0 / next 10.x > > re is cc'd on the original bugzilla issue If you would like to do so, please go ahead. Best, Conrad ___ 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: r301453 - in head/sys: arm/arm arm64/arm64 dev/fdt dev/gpio dev/iicbus dev/ofw dev/pci dev/vnic kern mips/mips sys
On 07/20/16 04:28, Michal Meloun wrote: Dne 19.07.2016 v 17:06 Nathan Whitehorn napsal(a): On 07/19/16 04:13, Michal Meloun wrote: Dne 19.07.2016 v 2:11 Nathan Whitehorn napsal(a): Hi Nathan, I’m afraid that skra is on vacation, for next 2 weeks (at minimum), so please don’t expect quick response. Could you please describe what this change is in more detail? Short description is appended. It breaks a lot of encapsulations we have worked very hard to maintain, moves ARM code into MI parts of the kernel, and the OFW parts violate IEEE 1275 (the Open Firmware standard). In particular, there is no guarantee that the interrupts for a newbus (or OF) device are encoded in a property called "interrupts" (or, indeed, in any property at all) on that node and there are many, many device trees where that is not the case (e.g. ones with interrupt maps, as well as Apple hardware). By putting that knowledge into the OF root bus device, which we have tried to keep it out of, this enforces a standard that doesn't actually exist. Imho, this patch doesn’t change anything in this area. Only handling of “interrupts” property is changed, all other cases are unchanged (I hope). Also, INTRNG code is currently shared by ARM, ARM64 and MIPS. But "interrupts" isn't a generic part of OF. This makes it one, incorrectly. How? Can you be little more exact ? Because it puts knowledge into ofwbus that expects that children at arbitrary levels of nesting have interrupts defined by an "interrupts" property. You could patch this through on sub-devices, of course, but that's already done correctly by the existing ofw_bus_map_intr() code in a much more robust way that doesn't involve trying to guess how sub-buses and devices have chosen to allocate resources. Why reinvent the wheel all the way through the bus hierarchy? I'm hesitant to ask for reversion on something that landed 6 weeks ago without me noticing, but this needs a lot more architectural work before any parts of the kernel should use it. -Nathan I think that it’s too late. This patch series consist of r301451 (https://reviews.freebsd.org/D6632), r301453, r301539 and 301543. And new GPIO interrupts are currently used (by in tree drivers or in development trees). Well, then we need in-place rearchitecture. The root of problem is that standard way of delivering interrupt resource to consumer driver doesn’t works in OFW world. So we have some fact: - the format of interrupt property is dependent of interrupt controller and only interrupt controller can parse it. - the interrupt property can have more data than just interrupt number. - single interrupt controller must be able to handle multiple format of interrupt description. In pre-patchset era, simplebus enumerates children and attempts to set memory and interrupts to resource list for them. But the interrupt controllers are not yet populated so nobody can parse interrupt property. Moreover, in all cases (parsed or not), we cannot store complete interrupt description into resource list. We have done this for many years on PowerPC and sparc64 with delayed configuration of interrupts and a look-up table. This handles complicated bus configurations better than this code and requires no changes outside of a few MD files. That is why the (now partially duplicated) OFW_BUS_MAP_INTR() function exists. That one also has the benefit of still working when used in conjunction with, e.g., devices with an interrupt-map-mask property. The patch simply postpones reading of interrupt property to bus_alloc_resource() (called by consumer driver) time. Due to this, we can: - parse interrupt property. The interrupt driver must exist at this time. This only works with some types of interrupt properties, not all, and breaks if the interrupt driver hasn't attached yet (which it can't be guaranteed to -- some PPC systems have interrupt drivers that live on the PCI bus, for example). How you can allocate (and reserve it in rman) interrupt if is not mapped (so you have not real IRQ number for it). Just for notice - multiple virtual IRQs can be mapped into single real IRQ. The core idea is to think of the full interrupt specifier -- the interrupt parent and the full byte string in the device tree -- as the IRQ rather than the interrupt pin on some chip (which is usually, but not always, the first word in that byte string). The "virtual" IRQ number is just a compression of that longer piece of data, which usually can't fit in an rman resource. There is no need to actually activate those interrupts before interrupts are enabled, so you can just cache them in a table until the end of device probing, which lets you break circular dependency loops between bus and interrupt topology. So long as you keep track of your mapping and the same (parent, interrupt specifier) parent always gives the same virtual IRQ, there is no way in this system to map multiple active IR
svn commit: r303094 - head/usr.bin
Author: ache Date: Wed Jul 20 15:59:37 2016 New Revision: 303094 URL: https://svnweb.freebsd.org/changeset/base/303094 Log: Continuation lines with comments badly affects gprof, it is excluded from build on amd64 f.e. Modified: head/usr.bin/Makefile Modified: head/usr.bin/Makefile == --- head/usr.bin/Makefile Wed Jul 20 15:07:52 2016(r303093) +++ head/usr.bin/Makefile Wed Jul 20 15:59:37 2016(r303094) @@ -270,8 +270,9 @@ SUBDIR.${MK_TOOLCHAIN}+=ctags SUBDIR.${MK_TOOLCHAIN}+= cxxfilt SUBDIR.${MK_TOOLCHAIN}+= elfcopy SUBDIR.${MK_TOOLCHAIN}+= file2c -.if ${MACHINE_ARCH} != "aarch64" && \ # ARM64TODO gprof does not build -${MACHINE_CPUARCH} != "riscv" # RISCVTODO gprof does not build +# ARM64TODO gprof does not build +# RISCVTODO gprof does not build +.if ${MACHINE_ARCH} != "aarch64" && ${MACHINE_CPUARCH} != "riscv" SUBDIR.${MK_TOOLCHAIN}+= gprof .endif SUBDIR.${MK_TOOLCHAIN}+= indent ___ 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: r303037 - head/sys/kern
Randall, I have just tested and r303037 brings the TCP panic back. I got two crashes during 2.5 hours. In your email [1] you are right, there is regression that (-1) return value is lost. This problem was worked on in the PR 210884, and we were very close to commiting the fix. The whole 11.0-RELEASE cycle strongly depends on this change. We don't want to release with TCP panic, and of course we want the regression described in 210884 to be fixed. Your backout mixed with extra code really made things messy. Since I don't want to go with commit war, on behalf of RE we are asking for explicit agreement to back out r303037. Then we will proceed with latest patch from 210884. Is that okay? [1] https://lists.freebsd.org/pipermail/svn-src-head/2016-July/089313.html On Wed, Jul 20, 2016 at 03:33:37PM +0200, Randall Stewart wrote: R> Gleb R> R> I wish you would have responded earlier.. I am more than glad to hand R> off all kern_timeout.c to you… please take it commit what you want to R> it and have it. I hate the code and I dislike having to touch it. R> R> Its yours.. I can assure you I will not touch it again. R> R> R R> > On Jul 20, 2016, at 8:53 AM, Gleb Smirnoff wrote: R> > R> > On Tue, Jul 19, 2016 at 06:31:19PM +, Randall Stewart wrote: R> > R> Author: rrs R> > R> Date: Tue Jul 19 18:31:19 2016 R> > R> New Revision: 303037 R> > R> URL: https://svnweb.freebsd.org/changeset/base/303037 R> > R> R> > R> Log: R> > R> This reverts out Gleb's changes and adds three small R> > R> fixes that I think closes up the races Gleb was R> > R> looking for. This is running quite nicely in Netflix and R> > R> now no longer causes TCP-tcb leaks. R> > R> R> > R> Differential Revision:7135 R> > R> > Just to notice that I am completely pissed of by this commit R> > war, that you started. R> > R> > I've been testing my changes properly, I gave people time to R> > review my changes. You didn't. R> > R> > From your explanation in other emails I see that you've been R> > testing your changes with a version of FreeBSD that is a heavily R> > modified FreeBSD 10, not 11. R> > R> > The new code you mixed with revert of mine, doesn't fix the R> > problem observed. It fixes another problem that you imagined, R> > which might exist, but isn't observed. We already discussed that R> > and you didn't prove it wrong. R> > R> > Your change doesn't even revert my change completely. R> > R> > -- R> > Totus tuus, Glebius. R> R> R> Randall Stewart R> r...@netflix.com R> 803-317-4952 R> R> R> R> R> R> -- Totus tuus, Glebius. ___ 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: r303095 - stable/10/usr.bin/mkuzip
Author: sobomax Date: Wed Jul 20 16:36:17 2016 New Revision: 303095 URL: https://svnweb.freebsd.org/changeset/base/303095 Log: MFC: merge in all new features and improvements into mkuzip(8) from current, which includes: o LZMA compression; o block de-duplication; o performance improvements; o multi-thread support. This includes the following revisions: r295943,r296626,r296628,r296810,r298504,r298505,r298577 Suggested by: emaste Added: stable/10/usr.bin/mkuzip/mkuz_blk.c (contents, props changed) stable/10/usr.bin/mkuzip/mkuz_blk.h (contents, props changed) stable/10/usr.bin/mkuzip/mkuz_blk_chain.h (contents, props changed) stable/10/usr.bin/mkuzip/mkuz_blockcache.c (contents, props changed) stable/10/usr.bin/mkuzip/mkuz_blockcache.h (contents, props changed) stable/10/usr.bin/mkuzip/mkuz_cfg.h (contents, props changed) stable/10/usr.bin/mkuzip/mkuz_cloop.h (contents, props changed) stable/10/usr.bin/mkuzip/mkuz_conveyor.c (contents, props changed) stable/10/usr.bin/mkuzip/mkuz_conveyor.h (contents, props changed) stable/10/usr.bin/mkuzip/mkuz_format.h (contents, props changed) stable/10/usr.bin/mkuzip/mkuz_fqueue.c (contents, props changed) stable/10/usr.bin/mkuzip/mkuz_fqueue.h (contents, props changed) stable/10/usr.bin/mkuzip/mkuz_lzma.c (contents, props changed) stable/10/usr.bin/mkuzip/mkuz_lzma.h (contents, props changed) stable/10/usr.bin/mkuzip/mkuz_time.c (contents, props changed) stable/10/usr.bin/mkuzip/mkuz_time.h (contents, props changed) stable/10/usr.bin/mkuzip/mkuz_zlib.c (contents, props changed) stable/10/usr.bin/mkuzip/mkuz_zlib.h (contents, props changed) stable/10/usr.bin/mkuzip/mkuzip.h (contents, props changed) Modified: stable/10/usr.bin/mkuzip/Makefile stable/10/usr.bin/mkuzip/mkuzip.8 stable/10/usr.bin/mkuzip/mkuzip.c Modified: stable/10/usr.bin/mkuzip/Makefile == --- stable/10/usr.bin/mkuzip/Makefile Wed Jul 20 15:59:37 2016 (r303094) +++ stable/10/usr.bin/mkuzip/Makefile Wed Jul 20 16:36:17 2016 (r303095) @@ -2,9 +2,10 @@ PROG= mkuzip MAN= mkuzip.8 +SRCS= mkuzip.c mkuz_blockcache.c mkuz_lzma.c mkuz_zlib.c mkuz_conveyor.c \ + mkuz_blk.c mkuz_fqueue.c mkuz_time.c -DPADD= ${LIBZ} -LDADD= -lz - +DPADD= ${LIBZ} ${LIBMD} ${LIBLZMA} ${LIBPTHREAD} +LDADD= -lz -lmd -llzma -lpthread .include Added: stable/10/usr.bin/mkuzip/mkuz_blk.c == --- /dev/null 00:00:00 1970 (empty, because file is newly added) +++ stable/10/usr.bin/mkuzip/mkuz_blk.c Wed Jul 20 16:36:17 2016 (r303095) @@ -0,0 +1,45 @@ +/* + * Copyright (c) 2004-2016 Maxim Sobolev + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + *notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + *notice, this list of conditions and the following disclaimer in the + *documentation and/or other materials provided with the distribution. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + */ + +#include +__FBSDID("$FreeBSD$"); + +#include +#include + +#include "mkuzip.h" +#include "mkuz_blk.h" + +struct mkuz_blk * +mkuz_blk_ctor(size_t blen) +{ +struct mkuz_blk *rval; + +rval = mkuz_safe_zmalloc(sizeof(struct mkuz_blk) + blen); +rval->alen = blen; +rval->br_offset = OFFSET_UNDEF; +return (rval); +} Added: stable/10/usr.bin/mkuzip/mkuz_blk.h == --- /dev/null 00:00:00 1970 (empty, because file is newly added) +++ stable/10/usr.bin/mkuzip/mkuz_blk.h Wed Jul 20 16:36:17 2016 (r303095) @@ -0,0 +1,48 @@ +/* + * Copyright (c) 2004-2016 Maxim Sobolev + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following
Re: svn commit: r303037 - head/sys/kern
You are most welcome to backout anything you like.. as far as I am concerned you own the code.. R > On Jul 20, 2016, at 6:35 PM, Gleb Smirnoff wrote: > > Randall, > > I have just tested and r303037 brings the TCP panic back. I > got two crashes during 2.5 hours. > > In your email [1] you are right, there is regression that (-1) > return value is lost. This problem was worked on in the PR 210884, > and we were very close to commiting the fix. > > The whole 11.0-RELEASE cycle strongly depends on this change. We > don't want to release with TCP panic, and of course we want the > regression described in 210884 to be fixed. > > Your backout mixed with extra code really made things messy. Since > I don't want to go with commit war, on behalf of RE we are asking > for explicit agreement to back out r303037. Then we will proceed with > latest patch from 210884. Is that okay? > > [1] https://lists.freebsd.org/pipermail/svn-src-head/2016-July/089313.html > > On Wed, Jul 20, 2016 at 03:33:37PM +0200, Randall Stewart wrote: > R> Gleb > R> > R> I wish you would have responded earlier.. I am more than glad to hand > R> off all kern_timeout.c to you… please take it commit what you want to > R> it and have it. I hate the code and I dislike having to touch it. > R> > R> Its yours.. I can assure you I will not touch it again. > R> > R> R > R> > On Jul 20, 2016, at 8:53 AM, Gleb Smirnoff wrote: > R> > > R> > On Tue, Jul 19, 2016 at 06:31:19PM +, Randall Stewart wrote: > R> > R> Author: rrs > R> > R> Date: Tue Jul 19 18:31:19 2016 > R> > R> New Revision: 303037 > R> > R> URL: https://svnweb.freebsd.org/changeset/base/303037 > R> > R> > R> > R> Log: > R> > R> This reverts out Gleb's changes and adds three small > R> > R> fixes that I think closes up the races Gleb was > R> > R> looking for. This is running quite nicely in Netflix and > R> > R> now no longer causes TCP-tcb leaks. > R> > R> > R> > R> Differential Revision: 7135 > R> > > R> > Just to notice that I am completely pissed of by this commit > R> > war, that you started. > R> > > R> > I've been testing my changes properly, I gave people time to > R> > review my changes. You didn't. > R> > > R> > From your explanation in other emails I see that you've been > R> > testing your changes with a version of FreeBSD that is a heavily > R> > modified FreeBSD 10, not 11. > R> > > R> > The new code you mixed with revert of mine, doesn't fix the > R> > problem observed. It fixes another problem that you imagined, > R> > which might exist, but isn't observed. We already discussed that > R> > and you didn't prove it wrong. > R> > > R> > Your change doesn't even revert my change completely. > R> > > R> > -- > R> > Totus tuus, Glebius. > R> > R> > R> Randall Stewart > R> r...@netflix.com > R> 803-317-4952 > R> > R> > R> > R> > R> > R> > > -- > Totus tuus, Glebius. Randall Stewart r...@netflix.com 803-317-4952 ___ 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: r303096 - head/sys/kern
Author: glebius Date: Wed Jul 20 16:44:22 2016 New Revision: 303096 URL: https://svnweb.freebsd.org/changeset/base/303096 Log: Revert r303037. It re-introduces the panic with TCP timers. Agreed by:rrs, re (gjb) Modified: head/sys/kern/kern_timeout.c Modified: head/sys/kern/kern_timeout.c == --- head/sys/kern/kern_timeout.cWed Jul 20 16:36:17 2016 (r303095) +++ head/sys/kern/kern_timeout.cWed Jul 20 16:44:22 2016 (r303096) @@ -1050,7 +1050,7 @@ callout_reset_sbt_on(struct callout *c, */ if (c->c_lock != NULL && !cc_exec_cancel(cc, direct)) cancelled = cc_exec_cancel(cc, direct) = true; - if (cc_exec_waiting(cc, direct) || cc_exec_drain(cc, direct)) { + if (cc_exec_waiting(cc, direct)) { /* * Someone has called callout_drain to kill this * callout. Don't reschedule. @@ -1166,7 +1166,7 @@ _callout_stop_safe(struct callout *c, in struct callout_cpu *cc, *old_cc; struct lock_class *class; int direct, sq_locked, use_lock; - int not_on_a_list; + int cancelled, not_on_a_list; if ((flags & CS_DRAIN) != 0) WITNESS_WARN(WARN_GIANTOK | WARN_SLEEPOK, c->c_lock, @@ -1234,47 +1234,17 @@ again: panic("migration should not happen"); #endif } - if ((drain != NULL) && (c->c_iflags & CALLOUT_PENDING) && - (cc_exec_curr(cc, direct) != c)) { - /* -* This callout is executing and we are draining. -* The only way this can happen is if its also -* been rescheduled to run on one thread *and* asked to drain -* on this thread (at the same time it is waiting to execute). -*/ - if ((c->c_iflags & CALLOUT_PROCESSED) == 0) { - if (cc_exec_next(cc) == c) - cc_exec_next(cc) = LIST_NEXT(c, c_links.le); - LIST_REMOVE(c, c_links.le); - } else { - TAILQ_REMOVE(&cc->cc_expireq, c, c_links.tqe); - } - c->c_iflags &= ~CALLOUT_PENDING; - c->c_flags &= ~CALLOUT_ACTIVE; - } + /* -* If the callout isn't pending, it's not on the queue, so -* don't attempt to remove it from the queue. We can try to -* stop it by other means however. +* If the callout is running, try to stop it or drain it. */ - if (!(c->c_iflags & CALLOUT_PENDING)) { + if (cc_exec_curr(cc, direct) == c) { /* -* If it wasn't on the queue and it isn't the current -* callout, then we can't stop it, so just bail. -* It probably has already been run (if locking -* is properly done). You could get here if the caller -* calls stop twice in a row for example. The second -* call would fall here without CALLOUT_ACTIVE set. +* Succeed we to stop it or not, we must clear the +* active flag - this is what API users expect. */ c->c_flags &= ~CALLOUT_ACTIVE; - if (cc_exec_curr(cc, direct) != c) { - CTR3(KTR_CALLOUT, "failed to stop %p func %p arg %p", - c, c->c_func, c->c_arg); - CC_UNLOCK(cc); - if (sq_locked) - sleepq_release(&cc_exec_waiting(cc, direct)); - return (-1); - } + if ((flags & CS_DRAIN) != 0) { /* * The current callout is running (or just @@ -1308,6 +1278,7 @@ again: old_cc = cc; goto again; } + /* * Migration could be cancelled here, but * as long as it is still not sure when it @@ -1391,8 +1362,6 @@ again: cc_exec_drain(cc, direct) = drain; } CC_UNLOCK(cc); - if (drain) - return(0); return ((flags & CS_EXECUTING) != 0); } CTR3(KTR_CALLOUT, "failed to stop %p func %p arg %p", @@ -1400,13 +1369,21 @@ again: if (drain) { cc_exec_drain(cc, direct) = drain; } - CC_UNLOCK(cc); KASSERT(!sq_locked, ("sleepqueue chain still locked")); - return (0); - } + cancelled = ((flags &
svn commit: r303098 - head/sys/kern
Author: glebius Date: Wed Jul 20 16:48:25 2016 New Revision: 303098 URL: https://svnweb.freebsd.org/changeset/base/303098 Log: Redo the r302894: the very new value for a non-scheduled callout is -1. This was recently added in r290664. Noticed by: hselasky Tested by:Larry Rosenman PR: 210884 Modified: head/sys/kern/kern_timeout.c Modified: head/sys/kern/kern_timeout.c == --- head/sys/kern/kern_timeout.cWed Jul 20 16:46:13 2016 (r303097) +++ head/sys/kern/kern_timeout.cWed Jul 20 16:48:25 2016 (r303098) @@ -1380,8 +1380,14 @@ again: if ((c->c_iflags & CALLOUT_PENDING) == 0) { CTR3(KTR_CALLOUT, "failed to stop %p func %p arg %p", c, c->c_func, c->c_arg); + /* +* For not scheduled and not executing callout return +* negative value. +*/ + if (cc_exec_curr(cc, direct) != c) + cancelled = -1; CC_UNLOCK(cc); - return (0); + return (cancelled); } c->c_iflags &= ~CALLOUT_PENDING; ___ 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: r303037 - head/sys/kern
On 07/20/16 18:35, Gleb Smirnoff wrote: Randall, I have just tested and r303037 brings the TCP panic back. I got two crashes during 2.5 hours. Hi Gleb, There is more than one way to fix the TCP panic. You don't _need_ to change the return values of callout_stop(). You _can_ make a new function for this, which doesn't cause problems in case anyone out there is depending on the current callout_stop() behaviour. In your email [1] you are right, there is regression that (-1) return value is lost. This problem was worked on in the PR 210884, and we were very close to commiting the fix. The whole 11.0-RELEASE cycle strongly depends on this change. We don't want to release with TCP panic, and of course we want the regression described in 210884 to be fixed. Your backout mixed with extra code really made things messy. Since I don't want to go with commit war, on behalf of RE we are asking for explicit agreement to back out r303037. Then we will proceed with latest patch from 210884. Is that okay? I think Randall is right trying to preserve the callout API. My wish Gleb, is that you add a new function to handle the new behaviour. [1] https://lists.freebsd.org/pipermail/svn-src-head/2016-July/089313.html --HPS ___ 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: r303097 - head
Author: nwhitehorn Date: Wed Jul 20 16:46:13 2016 New Revision: 303097 URL: https://svnweb.freebsd.org/changeset/base/303097 Log: Add myself to MAINTAINERS. Modified: head/MAINTAINERS Modified: head/MAINTAINERS == --- head/MAINTAINERSWed Jul 20 16:44:22 2016(r303096) +++ head/MAINTAINERSWed Jul 20 16:46:13 2016(r303097) @@ -105,3 +105,4 @@ vmm(4) neel,grehan Pre-commit review re autofs(5) trasz Pre-commit review recommended. iscsi(4) trasz Pre-commit review recommended. rctl(8)trasz Pre-commit review recommended. +sys/dev/ofwnwhitehorn Pre-commit review recommended. ___ 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: r303099 - head/sys/kern
Author: cem Date: Wed Jul 20 16:59:36 2016 New Revision: 303099 URL: https://svnweb.freebsd.org/changeset/base/303099 Log: Extend ELF coredump to support more than 65535 segments The ELF e_phnum field is only 16 bits wide. To support more than 65535 segments (program headers), Sun's "Linker and Libraries Guide" table 7-7 (or 12-7, depending on document version) prescribes a special first section header where sh_info represents the real number of program headers. Test code to follow, when it is ready. Reference:http://docs.oracle.com/cd/E18752_01/pdf/817-1984.pdf Reviewed by: emaste, markj Sponsored by: EMC / Isilon Storage Division Differential Revision:https://reviews.freebsd.org/D7255 Modified: head/sys/kern/imgact_elf.c Modified: head/sys/kern/imgact_elf.c == --- head/sys/kern/imgact_elf.c Wed Jul 20 16:48:25 2016(r303098) +++ head/sys/kern/imgact_elf.c Wed Jul 20 16:59:36 2016(r303099) @@ -1323,6 +1323,8 @@ __elfN(coredump)(struct thread *td, stru * Collect info about the core file header area. */ hdrsize = sizeof(Elf_Ehdr) + sizeof(Elf_Phdr) * (1 + seginfo.count); + if (seginfo.count + 1 >= PN_XNUM) + hdrsize += sizeof(Elf_Shdr); __elfN(prepare_notes)(td, ¬elst, ¬esz); coresize = round_page(hdrsize + notesz) + seginfo.size; @@ -1618,10 +1620,10 @@ __elfN(puthdr)(struct thread *td, void * { Elf_Ehdr *ehdr; Elf_Phdr *phdr; + Elf_Shdr *shdr; struct phdr_closure phc; ehdr = (Elf_Ehdr *)hdr; - phdr = (Elf_Phdr *)((char *)hdr + sizeof(Elf_Ehdr)); ehdr->e_ident[EI_MAG0] = ELFMAG0; ehdr->e_ident[EI_MAG1] = ELFMAG1; @@ -1645,14 +1647,43 @@ __elfN(puthdr)(struct thread *td, void * ehdr->e_flags = 0; ehdr->e_ehsize = sizeof(Elf_Ehdr); ehdr->e_phentsize = sizeof(Elf_Phdr); - ehdr->e_phnum = numsegs + 1; ehdr->e_shentsize = sizeof(Elf_Shdr); - ehdr->e_shnum = 0; ehdr->e_shstrndx = SHN_UNDEF; + if (numsegs + 1 < PN_XNUM) { + ehdr->e_phnum = numsegs + 1; + ehdr->e_shnum = 0; + } else { + ehdr->e_phnum = PN_XNUM; + ehdr->e_shnum = 1; + + ehdr->e_shoff = ehdr->e_phoff + + (numsegs + 1) * ehdr->e_phentsize; + KASSERT(ehdr->e_shoff == hdrsize - sizeof(Elf_Shdr), + ("e_shoff: %zu, hdrsize - shdr: %zu", +ehdr->e_shoff, hdrsize - sizeof(Elf_Shdr))); + + shdr = (Elf_Shdr *)((char *)hdr + ehdr->e_shoff); + memset(shdr, 0, sizeof(*shdr)); + /* +* A special first section is used to hold large segment and +* section counts. This was proposed by Sun Microsystems in +* Solaris and has been adopted by Linux; the standard ELF +* tools are already familiar with the technique. +* +* See table 7-7 of the Solaris "Linker and Libraries Guide" +* (or 12-7 depending on the version of the document) for more +* details. +*/ + shdr->sh_type = SHT_NULL; + shdr->sh_size = ehdr->e_shnum; + shdr->sh_link = ehdr->e_shstrndx; + shdr->sh_info = numsegs + 1; + } /* * Fill in the program header entries. */ + phdr = (Elf_Phdr *)((char *)hdr + ehdr->e_phoff); /* The note segement. */ phdr->p_type = PT_NOTE; ___ 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: r303100 - head/sys/dev/uart
Author: andrew Date: Wed Jul 20 17:19:47 2016 New Revision: 303100 URL: https://svnweb.freebsd.org/changeset/base/303100 Log: We will be switching to a new arm64 uart cpu driver that handles both FDT and ACPI. As such pull out what will be the common parts of the FDT cpu detection to a new function that can be shared between them. Reviewed by: manu Obtained from:ABT Systems Ltd MFC after:1 month Sponsored by: The FreeBSD Foundation Differential Revision:https://reviews.freebsd.org/D7262 Modified: head/sys/dev/uart/uart_bus_fdt.c head/sys/dev/uart/uart_cpu_fdt.c head/sys/dev/uart/uart_cpu_fdt.h Modified: head/sys/dev/uart/uart_bus_fdt.c == --- head/sys/dev/uart/uart_bus_fdt.cWed Jul 20 16:59:36 2016 (r303099) +++ head/sys/dev/uart/uart_bus_fdt.cWed Jul 20 17:19:47 2016 (r303100) @@ -105,6 +105,129 @@ uart_fdt_find_device(device_t dev) } static int +phandle_chosen_propdev(phandle_t chosen, const char *name, phandle_t *node) +{ + char buf[64]; + + if (OF_getprop(chosen, name, buf, sizeof(buf)) <= 0) + return (ENXIO); + if ((*node = OF_finddevice(buf)) == -1) + return (ENXIO); + + return (0); +} + +static const struct ofw_compat_data * +uart_fdt_find_compatible(phandle_t node, const struct ofw_compat_data *cd) +{ + const struct ofw_compat_data *ocd; + + for (ocd = cd; ocd->ocd_str != NULL; ocd++) { + if (fdt_is_compatible(node, ocd->ocd_str)) + return (ocd); + } + return (NULL); +} + +static uintptr_t +uart_fdt_find_by_node(phandle_t node, int class_list) +{ + struct ofw_compat_data **cd; + const struct ofw_compat_data *ocd; + + if (class_list) { + SET_FOREACH(cd, uart_fdt_class_set) { + ocd = uart_fdt_find_compatible(node, *cd); + if ((ocd != NULL) && (ocd->ocd_data != 0)) + return (ocd->ocd_data); + } + } else { + SET_FOREACH(cd, uart_fdt_class_and_device_set) { + ocd = uart_fdt_find_compatible(node, *cd); + if ((ocd != NULL) && (ocd->ocd_data != 0)) + return (ocd->ocd_data); + } + } + + return (0); +} + +int +uart_cpu_fdt_probe(struct uart_class **classp, bus_space_tag_t *bst, +bus_space_handle_t *bsh, int *baud, u_int *rclk, u_int *shiftp) +{ + const char *propnames[] = {"stdout-path", "linux,stdout-path", "stdout", + "stdin-path", "stdin", NULL}; + const char **name; + struct uart_class *class; + phandle_t node, chosen; + pcell_t br, clk, shift; + char *cp; + int err; + + /* Has the user forced a specific device node? */ + cp = kern_getenv("hw.fdt.console"); + if (cp == NULL) { + /* +* Retrieve /chosen/std{in,out}. +*/ + node = -1; + if ((chosen = OF_finddevice("/chosen")) != -1) { + for (name = propnames; *name != NULL; name++) { + if (phandle_chosen_propdev(chosen, *name, + &node) == 0) + break; + } + } + if (chosen == -1 || *name == NULL) + node = OF_finddevice("serial0"); /* Last ditch */ + } else { + node = OF_finddevice(cp); + } + + if (node == -1) + return (ENXIO); + + /* +* Check old style of UART definition first. Unfortunately, the common +* FDT processing is not possible if we have clock, power domains and +* pinmux stuff. +*/ + class = (struct uart_class *)uart_fdt_find_by_node(node, 0); + if (class != NULL) { + if ((err = uart_fdt_get_clock(node, &clk)) != 0) + return (err); + } else { + /* Check class only linker set */ + class = + (struct uart_class *)uart_fdt_find_by_node(node, 1); + if (class == NULL) + return (ENXIO); + clk = 0; + } + + /* +* Retrieve serial attributes. +*/ + if (uart_fdt_get_shift(node, &shift) != 0) + shift = uart_getregshift(class); + + if (OF_getencprop(node, "current-speed", &br, sizeof(br)) <= 0) + br = 0; + + err = OF_decode_addr(node, 0, bst, bsh, NULL); + if (err != 0) + return (err); + + *classp = class; + *baud = br; + *rclk = clk; + *shiftp = shift; + + return (0); +} + +static int uart_fdt_probe(device_t dev) { struct uart_softc *sc; Modified: head/sys/dev/uart/ua
svn commit: r303101 - head/sys/vm
Author: alc Date: Wed Jul 20 17:20:22 2016 New Revision: 303101 URL: https://svnweb.freebsd.org/changeset/base/303101 Log: Add a comment describing the 'fast path' that was introduced in r270011. Reviewed by: kib MFC after:3 days Sponsored by: EMC / Isilon Storage Division Modified: head/sys/vm/vm_fault.c Modified: head/sys/vm/vm_fault.c == --- head/sys/vm/vm_fault.c Wed Jul 20 17:19:47 2016(r303100) +++ head/sys/vm/vm_fault.c Wed Jul 20 17:20:22 2016(r303101) @@ -354,6 +354,15 @@ RetryFault:; KASSERT((fault_flags & VM_FAULT_WIRE) == 0, ("!wired && VM_FAULT_WIRE")); + /* +* Try to avoid lock contention on the top-level object through +* special-case handling of some types of page faults, specifically, +* those that are both (1) mapping an existing page from the top- +* level object and (2) not having to mark that object as containing +* dirty pages. Under these conditions, a read lock on the top-level +* object suffices, allowing multiple page faults of a similar type to +* run in parallel on the same top-level object. +*/ if (fs.vp == NULL /* avoid locked vnode leak */ && (fault_flags & (VM_FAULT_WIRE | VM_FAULT_DIRTY)) == 0 && /* avoid calling vm_object_set_writeable_dirty() */ ___ 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: r303099 - head/sys/kern
This broke my build: --- imgact_elf32.o --- In file included from /usr/src/sys/kern/imgact_elf32.c:31: /usr/src/sys/kern/imgact_elf.c:1663:8: error: format specifies type 'size_t' (aka 'unsigned long') but the argument has type 'Elf32_Off' (aka 'unsigned int') [-Werror,-Wformat] ehdr->e_shoff, hdrsize - sizeof(Elf_Shdr))); ^ /usr/src/sys/sys/systm.h:86:17: note: expanded from macro 'KASSERT' kassert_panic msg; \ ^~~ On 2016-07-20 11:59, Conrad E. Meyer wrote: Author: cem Date: Wed Jul 20 16:59:36 2016 New Revision: 303099 URL: https://svnweb.freebsd.org/changeset/base/303099 Log: Extend ELF coredump to support more than 65535 segments The ELF e_phnum field is only 16 bits wide. To support more than 65535 segments (program headers), Sun's "Linker and Libraries Guide" table 7-7 (or 12-7, depending on document version) prescribes a special first section header where sh_info represents the real number of program headers. Test code to follow, when it is ready. Reference:http://docs.oracle.com/cd/E18752_01/pdf/817-1984.pdf Reviewed by: emaste, markj Sponsored by: EMC / Isilon Storage Division Differential Revision:https://reviews.freebsd.org/D7255 Modified: head/sys/kern/imgact_elf.c Modified: head/sys/kern/imgact_elf.c == --- head/sys/kern/imgact_elf.c Wed Jul 20 16:48:25 2016(r303098) +++ head/sys/kern/imgact_elf.c Wed Jul 20 16:59:36 2016(r303099) @@ -1323,6 +1323,8 @@ __elfN(coredump)(struct thread *td, stru * Collect info about the core file header area. */ hdrsize = sizeof(Elf_Ehdr) + sizeof(Elf_Phdr) * (1 + seginfo.count); + if (seginfo.count + 1 >= PN_XNUM) + hdrsize += sizeof(Elf_Shdr); __elfN(prepare_notes)(td, ¬elst, ¬esz); coresize = round_page(hdrsize + notesz) + seginfo.size; @@ -1618,10 +1620,10 @@ __elfN(puthdr)(struct thread *td, void * { Elf_Ehdr *ehdr; Elf_Phdr *phdr; + Elf_Shdr *shdr; struct phdr_closure phc; ehdr = (Elf_Ehdr *)hdr; - phdr = (Elf_Phdr *)((char *)hdr + sizeof(Elf_Ehdr)); ehdr->e_ident[EI_MAG0] = ELFMAG0; ehdr->e_ident[EI_MAG1] = ELFMAG1; @@ -1645,14 +1647,43 @@ __elfN(puthdr)(struct thread *td, void * ehdr->e_flags = 0; ehdr->e_ehsize = sizeof(Elf_Ehdr); ehdr->e_phentsize = sizeof(Elf_Phdr); - ehdr->e_phnum = numsegs + 1; ehdr->e_shentsize = sizeof(Elf_Shdr); - ehdr->e_shnum = 0; ehdr->e_shstrndx = SHN_UNDEF; + if (numsegs + 1 < PN_XNUM) { + ehdr->e_phnum = numsegs + 1; + ehdr->e_shnum = 0; + } else { + ehdr->e_phnum = PN_XNUM; + ehdr->e_shnum = 1; + + ehdr->e_shoff = ehdr->e_phoff + + (numsegs + 1) * ehdr->e_phentsize; + KASSERT(ehdr->e_shoff == hdrsize - sizeof(Elf_Shdr), + ("e_shoff: %zu, hdrsize - shdr: %zu", +ehdr->e_shoff, hdrsize - sizeof(Elf_Shdr))); + + shdr = (Elf_Shdr *)((char *)hdr + ehdr->e_shoff); + memset(shdr, 0, sizeof(*shdr)); + /* +* A special first section is used to hold large segment and +* section counts. This was proposed by Sun Microsystems in +* Solaris and has been adopted by Linux; the standard ELF +* tools are already familiar with the technique. +* +* See table 7-7 of the Solaris "Linker and Libraries Guide" +* (or 12-7 depending on the version of the document) for more +* details. +*/ + shdr->sh_type = SHT_NULL; + shdr->sh_size = ehdr->e_shnum; + shdr->sh_link = ehdr->e_shstrndx; + shdr->sh_info = numsegs + 1; + } /* * Fill in the program header entries. */ + phdr = (Elf_Phdr *)((char *)hdr + ehdr->e_phoff); /* The note segement. */ phdr->p_type = PT_NOTE; ___ 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" -- Larry Rosenman http://www.lerctr.org/~ler Phone: +1 214-642-9640 E-Mail: l...@lerctr.org US Mail: 17716 Limpia Crk, Round Rock, TX 78664-7281 ___ 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: r303102 - head/sys/conf
Author: andrew Date: Wed Jul 20 17:46:33 2016 New Revision: 303102 URL: https://svnweb.freebsd.org/changeset/base/303102 Log: Mark the Designware MMC and USB OTG drivers as FDT only. These are normally found on arm64 devices that use FDT. Obtained from:ABT Systems Ltd MFC after:1 month Sponsored by: The FreeBSD Foundation Modified: head/sys/conf/files.arm64 Modified: head/sys/conf/files.arm64 == --- head/sys/conf/files.arm64 Wed Jul 20 17:20:22 2016(r303101) +++ head/sys/conf/files.arm64 Wed Jul 20 17:46:33 2016(r303102) @@ -59,8 +59,8 @@ dev/acpica/acpi_if.m optionalacpi dev/ahci/ahci_generic.coptional ahci fdt dev/hwpmc/hwpmc_arm64.coptionalhwpmc dev/hwpmc/hwpmc_arm64_md.c optionalhwpmc -dev/mmc/host/dwmmc.c optionaldwmmc -dev/mmc/host/dwmmc_hisi.c optionaldwmmc soc_hisi_hi6220 +dev/mmc/host/dwmmc.c optionaldwmmc fdt +dev/mmc/host/dwmmc_hisi.c optionaldwmmc fdt soc_hisi_hi6220 dev/ofw/ofw_cpu.c optionalfdt dev/ofw/ofwpci.c optionalfdt pci dev/pci/pci_host_generic.c optionalpci fdt @@ -68,7 +68,7 @@ dev/psci/psci.c optionalpsci dev/psci/psci_arm64.S optionalpsci dev/uart/uart_cpu_fdt.coptionaluart fdt dev/uart/uart_dev_pl011.c optionaluart pl011 -dev/usb/controller/dwc_otg_hisi.c optional dwcotg soc_hisi_hi6220 +dev/usb/controller/dwc_otg_hisi.c optional dwcotg fdt soc_hisi_hi6220 dev/usb/controller/generic_ohci.c optional ohci fdt dev/usb/controller/generic_usb_if.m optional ohci fdt dev/vnic/mrml_bridge.c optionalvnic fdt ___ 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: r303103 - head
Author: brooks Date: Wed Jul 20 18:00:22 2016 New Revision: 303103 URL: https://svnweb.freebsd.org/changeset/base/303103 Log: Minor wording improvements to the note about pipe(2) removal. MFC after:5 days Sponsored by: DARPA, AFRL Modified: head/UPDATING Modified: head/UPDATING == --- head/UPDATING Wed Jul 20 17:46:33 2016(r303102) +++ head/UPDATING Wed Jul 20 18:00:22 2016(r303103) @@ -33,11 +33,12 @@ NOTE TO PEOPLE WHO THINK THAT FreeBSD 12 20160622: The libc stub for the pipe(2) system call has been replaced with - a wrapper which calls the pipe2(2) system call and the pipe(2) is now - only implemented by the kernels which include "options - COMPAT_FREEBSD10" in their config file (this is the default). - Users should ensure that this option is enabled in their kernel - or upgrade userspace to r302092 before upgrading their kernel. + a wrapper that calls the pipe2(2) system call and the pipe(2) + system call is now only implemented by the kernels that include + "options COMPAT_FREEBSD10" in their config file (this is the + default). Users should ensure that this option is enabled in + their kernel or upgrade userspace to r302092 before upgrading their + kernel. 20160527: CAM will now strip leading spaces from SCSI disks' serial numbers. ___ 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: r303104 - head/lib/libc/sys
Author: brooks Date: Wed Jul 20 18:02:07 2016 New Revision: 303104 URL: https://svnweb.freebsd.org/changeset/base/303104 Log: Update to reflect the fact that pipe() is a wrapper around the pipe2() system call. Reviewed by: jhb, wblock MFC after:5 days Sponsored by: DAPRA, AFRL Differential Revision:https://reviews.freebsd.org/D6948 Modified: head/lib/libc/sys/pipe.2 Modified: head/lib/libc/sys/pipe.2 == --- head/lib/libc/sys/pipe.2Wed Jul 20 18:00:22 2016(r303103) +++ head/lib/libc/sys/pipe.2Wed Jul 20 18:02:07 2016(r303104) @@ -28,7 +28,7 @@ .\" @(#)pipe.2 8.1 (Berkeley) 6/4/93 .\" $FreeBSD$ .\" -.Dd June 22, 2016 +.Dd July 20, 2016 .Dt PIPE 2 .Os .Sh NAME @@ -109,6 +109,18 @@ The bidirectional nature of this impleme portable to older systems, so it is recommended to use the convention for using the endpoints in the traditional manner when using a pipe in one direction. +.Sh IMPLEMENTATION NOTES +The +.Fn pipe +function calls the +.Fn pipe2 +system call. +As a result, system call traces such as those captured by +.Xr dtrace 1 +or +.Xr ktrace 1 +will show calls to +.Fn pipe2 . .Sh RETURN VALUES .Rv -std pipe .Sh ERRORS @@ -157,3 +169,10 @@ The .Fn pipe2 function appeared in .Fx 10.0 . +.Pp +The +.Fn pipe +function became a wrapper around +.Fn pipe2 +in +.Fx 11.0 . ___ 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: r303105 - head/sys/kern
Author: cem Date: Wed Jul 20 18:11:22 2016 New Revision: 303105 URL: https://svnweb.freebsd.org/changeset/base/303105 Log: Fix DEBUG build on 64-bit arch after r303099 Reported by: Larry Rosenman Modified: head/sys/kern/imgact_elf.c Modified: head/sys/kern/imgact_elf.c == --- head/sys/kern/imgact_elf.c Wed Jul 20 18:02:07 2016(r303104) +++ head/sys/kern/imgact_elf.c Wed Jul 20 18:11:22 2016(r303105) @@ -1660,7 +1660,7 @@ __elfN(puthdr)(struct thread *td, void * (numsegs + 1) * ehdr->e_phentsize; KASSERT(ehdr->e_shoff == hdrsize - sizeof(Elf_Shdr), ("e_shoff: %zu, hdrsize - shdr: %zu", -ehdr->e_shoff, hdrsize - sizeof(Elf_Shdr))); +(size_t)ehdr->e_shoff, hdrsize - sizeof(Elf_Shdr))); shdr = (Elf_Shdr *)((char *)hdr + ehdr->e_shoff); memset(shdr, 0, sizeof(*shdr)); ___ 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: r303099 - head/sys/kern
Mea culpa. Thanks for the report. Fixed in r303105. Best, Conrad On Wed, Jul 20, 2016 at 10:32 AM, Larry Rosenman wrote: > This broke my build: > --- imgact_elf32.o --- > In file included from /usr/src/sys/kern/imgact_elf32.c:31: > /usr/src/sys/kern/imgact_elf.c:1663:8: error: format specifies type 'size_t' > (aka 'unsigned long') but the argument has type 'Elf32_Off' (aka 'unsigned > int') [-Werror,-Wformat] > ehdr->e_shoff, hdrsize - sizeof(Elf_Shdr))); > ^ > /usr/src/sys/sys/systm.h:86:17: note: expanded from macro 'KASSERT' > kassert_panic msg; \ > ^~~ > > > > > On 2016-07-20 11:59, Conrad E. Meyer wrote: >> >> Author: cem >> Date: Wed Jul 20 16:59:36 2016 >> New Revision: 303099 >> URL: https://svnweb.freebsd.org/changeset/base/303099 >> >> Log: >> Extend ELF coredump to support more than 65535 segments >> >> The ELF e_phnum field is only 16 bits wide. To support more than >> 65535 segments >> (program headers), Sun's "Linker and Libraries Guide" table 7-7 (or >> 12-7, >> depending on document version) prescribes a special first section header >> where >> sh_info represents the real number of program headers. >> >> Test code to follow, when it is ready. >> >> Reference:http://docs.oracle.com/cd/E18752_01/pdf/817-1984.pdf >> >> Reviewed by: emaste, markj >> Sponsored by: EMC / Isilon Storage Division >> Differential Revision:https://reviews.freebsd.org/D7255 >> >> Modified: >> head/sys/kern/imgact_elf.c >> >> Modified: head/sys/kern/imgact_elf.c >> >> == >> --- head/sys/kern/imgact_elf.c Wed Jul 20 16:48:25 2016(r303098) >> +++ head/sys/kern/imgact_elf.c Wed Jul 20 16:59:36 2016(r303099) >> @@ -1323,6 +1323,8 @@ __elfN(coredump)(struct thread *td, stru >> * Collect info about the core file header area. >> */ >> hdrsize = sizeof(Elf_Ehdr) + sizeof(Elf_Phdr) * (1 + >> seginfo.count); >> + if (seginfo.count + 1 >= PN_XNUM) >> + hdrsize += sizeof(Elf_Shdr); >> __elfN(prepare_notes)(td, ¬elst, ¬esz); >> coresize = round_page(hdrsize + notesz) + seginfo.size; >> >> @@ -1618,10 +1620,10 @@ __elfN(puthdr)(struct thread *td, void * >> { >> Elf_Ehdr *ehdr; >> Elf_Phdr *phdr; >> + Elf_Shdr *shdr; >> struct phdr_closure phc; >> >> ehdr = (Elf_Ehdr *)hdr; >> - phdr = (Elf_Phdr *)((char *)hdr + sizeof(Elf_Ehdr)); >> >> ehdr->e_ident[EI_MAG0] = ELFMAG0; >> ehdr->e_ident[EI_MAG1] = ELFMAG1; >> @@ -1645,14 +1647,43 @@ __elfN(puthdr)(struct thread *td, void * >> ehdr->e_flags = 0; >> ehdr->e_ehsize = sizeof(Elf_Ehdr); >> ehdr->e_phentsize = sizeof(Elf_Phdr); >> - ehdr->e_phnum = numsegs + 1; >> ehdr->e_shentsize = sizeof(Elf_Shdr); >> - ehdr->e_shnum = 0; >> ehdr->e_shstrndx = SHN_UNDEF; >> + if (numsegs + 1 < PN_XNUM) { >> + ehdr->e_phnum = numsegs + 1; >> + ehdr->e_shnum = 0; >> + } else { >> + ehdr->e_phnum = PN_XNUM; >> + ehdr->e_shnum = 1; >> + >> + ehdr->e_shoff = ehdr->e_phoff + >> + (numsegs + 1) * ehdr->e_phentsize; >> + KASSERT(ehdr->e_shoff == hdrsize - sizeof(Elf_Shdr), >> + ("e_shoff: %zu, hdrsize - shdr: %zu", >> +ehdr->e_shoff, hdrsize - sizeof(Elf_Shdr))); >> + >> + shdr = (Elf_Shdr *)((char *)hdr + ehdr->e_shoff); >> + memset(shdr, 0, sizeof(*shdr)); >> + /* >> +* A special first section is used to hold large segment >> and >> +* section counts. This was proposed by Sun Microsystems >> in >> +* Solaris and has been adopted by Linux; the standard ELF >> +* tools are already familiar with the technique. >> +* >> +* See table 7-7 of the Solaris "Linker and Libraries >> Guide" >> +* (or 12-7 depending on the version of the document) for >> more >> +* details. >> +*/ >> + shdr->sh_type = SHT_NULL; >> + shdr->sh_size = ehdr->e_shnum; >> + shdr->sh_link = ehdr->e_shstrndx; >> + shdr->sh_info = numsegs + 1; >> + } >> >> /* >> * Fill in the program header entries. >> */ >> + phdr = (Elf_Phdr *)((char *)hdr + ehdr->e_phoff); >> >> /* The note segement. */ >> phdr->p_type = PT_NOTE; >> ___ >> 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" > > > -- > Larry Rosenman
svn commit: r303106 - head/lib/libc/sys
Author: zeising (doc,ports committer) Date: Wed Jul 20 18:16:58 2016 New Revision: 303106 URL: https://svnweb.freebsd.org/changeset/base/303106 Log: Change wording to use function rather than system call in the description as well. Reviewed by: brooks MFC after:5 days Modified: head/lib/libc/sys/pipe.2 Modified: head/lib/libc/sys/pipe.2 == --- head/lib/libc/sys/pipe.2Wed Jul 20 18:11:22 2016(r303105) +++ head/lib/libc/sys/pipe.2Wed Jul 20 18:16:58 2016(r303106) @@ -46,7 +46,7 @@ .Sh DESCRIPTION The .Fn pipe -system call +function creates a .Em pipe , which is an object allowing ___ 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: r303107 - in stable/10/sys/dev: e1000 ixgb ixgbe ixl
Author: sbruno Date: Wed Jul 20 18:26:48 2016 New Revision: 303107 URL: https://svnweb.freebsd.org/changeset/base/303107 Log: MFC - r302384 to 10-STABLE Do not initialize the adapter on MTU change when adapter status is down. This fixes long-standing problems when changing settings of the adapter. Discussed in: https://lists.freebsd.org/pipermail/freebsd-net/2016-June/045509.html Differential Revision:https://reviews.freebsd.org/D7254 Modified: stable/10/sys/dev/e1000/if_em.c stable/10/sys/dev/e1000/if_igb.c stable/10/sys/dev/e1000/if_lem.c stable/10/sys/dev/ixgb/if_ixgb.c stable/10/sys/dev/ixgbe/if_ix.c stable/10/sys/dev/ixgbe/if_ixv.c stable/10/sys/dev/ixl/if_ixl.c stable/10/sys/dev/ixl/if_ixlv.c Directory Properties: stable/10/ (props changed) Modified: stable/10/sys/dev/e1000/if_em.c == --- stable/10/sys/dev/e1000/if_em.c Wed Jul 20 18:16:58 2016 (r303106) +++ stable/10/sys/dev/e1000/if_em.c Wed Jul 20 18:26:48 2016 (r303107) @@ -1220,7 +1220,8 @@ em_ioctl(struct ifnet *ifp, u_long comma ifp->if_mtu = ifr->ifr_mtu; adapter->hw.mac.max_frame_size = ifp->if_mtu + ETHER_HDR_LEN + ETHER_CRC_LEN; - em_init_locked(adapter); + if (if_getdrvflags(ifp) & IFF_DRV_RUNNING) + em_init_locked(adapter); EM_CORE_UNLOCK(adapter); break; } Modified: stable/10/sys/dev/e1000/if_igb.c == --- stable/10/sys/dev/e1000/if_igb.cWed Jul 20 18:16:58 2016 (r303106) +++ stable/10/sys/dev/e1000/if_igb.cWed Jul 20 18:26:48 2016 (r303107) @@ -1090,7 +1090,8 @@ igb_ioctl(struct ifnet *ifp, u_long comm ifp->if_mtu = ifr->ifr_mtu; adapter->max_frame_size = ifp->if_mtu + ETHER_HDR_LEN + ETHER_CRC_LEN; - igb_init_locked(adapter); + if ((ifp->if_drv_flags & IFF_DRV_RUNNING)) + igb_init_locked(adapter); IGB_CORE_UNLOCK(adapter); break; } Modified: stable/10/sys/dev/e1000/if_lem.c == --- stable/10/sys/dev/e1000/if_lem.cWed Jul 20 18:16:58 2016 (r303106) +++ stable/10/sys/dev/e1000/if_lem.cWed Jul 20 18:26:48 2016 (r303107) @@ -1047,7 +1047,8 @@ lem_ioctl(struct ifnet *ifp, u_long comm ifp->if_mtu = ifr->ifr_mtu; adapter->max_frame_size = ifp->if_mtu + ETHER_HDR_LEN + ETHER_CRC_LEN; - lem_init_locked(adapter); + if ((if_getdrvflags(ifp) & IFF_DRV_RUNNING)) + lem_init_locked(adapter); EM_CORE_UNLOCK(adapter); break; } Modified: stable/10/sys/dev/ixgb/if_ixgb.c == --- stable/10/sys/dev/ixgb/if_ixgb.cWed Jul 20 18:16:58 2016 (r303106) +++ stable/10/sys/dev/ixgb/if_ixgb.cWed Jul 20 18:26:48 2016 (r303107) @@ -538,7 +538,8 @@ ixgb_ioctl(struct ifnet * ifp, IOCTL_CMD adapter->hw.max_frame_size = ifp->if_mtu + ETHER_HDR_LEN + ETHER_CRC_LEN; - ixgb_init_locked(adapter); + if (ifp->if_drv_flags & IFF_DRV_RUNNING) + ixgb_init_locked(adapter); IXGB_UNLOCK(adapter); } break; Modified: stable/10/sys/dev/ixgbe/if_ix.c == --- stable/10/sys/dev/ixgbe/if_ix.c Wed Jul 20 18:16:58 2016 (r303106) +++ stable/10/sys/dev/ixgbe/if_ix.c Wed Jul 20 18:26:48 2016 (r303107) @@ -900,7 +900,8 @@ ixgbe_ioctl(struct ifnet * ifp, u_long c ifp->if_mtu = ifr->ifr_mtu; adapter->max_frame_size = ifp->if_mtu + IXGBE_MTU_HDR; - ixgbe_init_locked(adapter); + if (ifp->if_drv_flags & IFF_DRV_RUNNING) + ixgbe_init_locked(adapter); #ifdef PCI_IOV ixgbe_recalculate_max_frame(adapter); #endif Modified: stable/10/sys/dev/ixgbe/if_ixv.c == --- stable/10/sys/dev/ixgbe/if_ixv.cWed Jul 20 18:16:58 2016 (r303106) +++ stable/10/sys/dev/ixgbe/if_ixv.cWed Jul 20 18:26:48 2016 (r303107) @@ -587,7 +587,8 @@ ixv_ioctl(struct ifnet * ifp, u_long com ifp->if_mtu = ifr->ifr_mtu; adapter->max_frame_size =
svn commit: r303109 - head/usr.sbin/crashinfo
Author: jhb Date: Wed Jul 20 18:41:47 2016 New Revision: 303109 URL: https://svnweb.freebsd.org/changeset/base/303109 Log: Update crashinfo to work with newer gdb from ports. If gdb from ports is installed, use it instead of the base system gdb to extract variables from a kernel. Note that base gdb and ports gdb do not support the same options for invoking a single command in batch mode, so a wrapper shell function is used. In addition, prefer kgdb from ports when generating a backtrace if present. PR: 193335 Reviewed by: emaste MFC after:1 week Differential Revision:https://reviews.freebsd.org/D7218 Modified: head/usr.sbin/crashinfo/crashinfo.sh Modified: head/usr.sbin/crashinfo/crashinfo.sh == --- head/usr.sbin/crashinfo/crashinfo.shWed Jul 20 18:27:42 2016 (r303108) +++ head/usr.sbin/crashinfo/crashinfo.shWed Jul 20 18:41:47 2016 (r303109) @@ -35,6 +35,22 @@ usage() exit 1 } +# Run a single gdb command against a kernel file in batch mode. +# The kernel file is specified as the first argument and the command +# is given in the remaining arguments. +gdb_command() +{ + local k + + k=$1 ; shift + + if [ -x /usr/local/bin/gdb ]; then + /usr/local/bin/gdb -batch -ex "$@" $k + else + echo -e "$@" | /usr/bin/gdb -x /dev/stdin -batch $k + fi +} + find_kernel() { local ivers k kvers @@ -55,8 +71,8 @@ find_kernel() # Look for a matching kernel version. for k in `sysctl -n kern.bootfile` $(ls -t /boot/*/kernel); do - kvers=$(echo 'printf " Version String: %s", version' | \ - gdb -x /dev/stdin -batch $k 2>/dev/null) + kvers=$(gdb_command $k 'printf " Version String: %s", version' \ +2>/dev/null) if [ "$ivers" = "$kvers" ]; then KERNEL=$k break @@ -151,11 +167,10 @@ echo "Writing crash summary to $FILE." umask 077 # Simulate uname -ostype=$(echo -e printf '"%s", ostype' | gdb -x /dev/stdin -batch $KERNEL) -osrelease=$(echo -e printf '"%s", osrelease' | gdb -x /dev/stdin -batch $KERNEL) -version=$(echo -e printf '"%s", version' | gdb -x /dev/stdin -batch $KERNEL | \ -tr '\t\n' ' ') -machine=$(echo -e printf '"%s", machine' | gdb -x /dev/stdin -batch $KERNEL) +ostype=$(gdb_command $KERNEL 'printf "%s", ostype') +osrelease=$(gdb_command $KERNEL 'printf "%s", osrelease') +version=$(gdb_command $KERNEL 'printf "%s", version' | tr '\t\n' ' ') +machine=$(gdb_command $KERNEL 'printf "%s", machine') exec > $FILE 2>&1 @@ -174,7 +189,11 @@ file=`mktemp /tmp/crashinfo.XX` if [ $? -eq 0 ]; then echo "bt" >> $file echo "quit" >> $file - kgdb $KERNEL $VMCORE < $file + if [ -x /usr/local/bin/kgdb ]; then + /usr/local/bin/kgdb $KERNEL $VMCORE < $file + else + kgdb $KERNEL $VMCORE < $file + fi rm -f $file echo fi ___ 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: r303043 - in head: share/man/man4 sys/dev/vt sys/dev/vt/hw/fb
On 7/20/16 8:20 AM, Conrad Meyer wrote: > On Wed, Jul 20, 2016 at 12:44 AM, Kubilay Kocak wrote: >> On 20/07/2016 5:20 AM, Conrad E. Meyer wrote: >>> Author: cem >>> Date: Tue Jul 19 19:20:47 2016 >>> New Revision: 303043 >>> URL: https://svnweb.freebsd.org/changeset/base/303043 >>> >>> Log: >>> Increase vt(4) framebuffer maximum size >>> >>> And rename "DEFAULT" constants to the more accurate "MAX." >>> >>> PR: 210382 >>> Submitted by: Felix >>> Reviewed by:wblock, cem >>> Tested by: Dave Cottlehuber >> >> Can MFC? >> >> There are high hopes for this coming out in 11.0 / next 10.x >> >> re is cc'd on the original bugzilla issue > > If you would like to do so, please go ahead. > > Best, > Conrad > I'll handle it after it's been in for a week. -- Regards, Bryan Drewery ___ 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: r303110 - head/sys/dev/e1000
Author: sbruno Date: Wed Jul 20 19:21:11 2016 New Revision: 303110 URL: https://svnweb.freebsd.org/changeset/base/303110 Log: Remove uneeded parens. MFC after:1 day Modified: head/sys/dev/e1000/if_igb.c head/sys/dev/e1000/if_lem.c Modified: head/sys/dev/e1000/if_igb.c == --- head/sys/dev/e1000/if_igb.c Wed Jul 20 18:41:47 2016(r303109) +++ head/sys/dev/e1000/if_igb.c Wed Jul 20 19:21:11 2016(r303110) @@ -1106,7 +1106,7 @@ igb_ioctl(struct ifnet *ifp, u_long comm ifp->if_mtu = ifr->ifr_mtu; adapter->max_frame_size = ifp->if_mtu + ETHER_HDR_LEN + ETHER_CRC_LEN; - if ((ifp->if_drv_flags & IFF_DRV_RUNNING)) + if (ifp->if_drv_flags & IFF_DRV_RUNNING) igb_init_locked(adapter); IGB_CORE_UNLOCK(adapter); break; Modified: head/sys/dev/e1000/if_lem.c == --- head/sys/dev/e1000/if_lem.c Wed Jul 20 18:41:47 2016(r303109) +++ head/sys/dev/e1000/if_lem.c Wed Jul 20 19:21:11 2016(r303110) @@ -1053,7 +1053,7 @@ lem_ioctl(if_t ifp, u_long command, cadd if_setmtu(ifp, ifr->ifr_mtu); adapter->max_frame_size = if_getmtu(ifp) + ETHER_HDR_LEN + ETHER_CRC_LEN; - if ((if_getdrvflags(ifp) & IFF_DRV_RUNNING)) + if (if_getdrvflags(ifp) & IFF_DRV_RUNNING) lem_init_locked(adapter); EM_CORE_UNLOCK(adapter); break; ___ 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: r302350 - in head: share/man/man9 sys/kern sys/sys
On 7/13/16 3:11 AM, Hans Petter Selasky wrote: > Hi, > > Can we agree on the following return values for callout_stop() and > callout_reset() and their variants: > > If the callout was scheduled regardless of being serviced or not, > .Dv CALLOUT_RET_CANCELLED > is returned. > If the callout was stopped and is still being serviced > .Dv CALLOUT_RET_DRAINING > is returned. > If the callout was stopped and is no longer being serviced > .Dv CALLOUT_RET_STOPPED > is returned. > > --HPS > I won't comment on the specific cases, but I do think we should move to using enums or macros for these return values and other parts of the KPI. The callout API is getting very abused lately, since r290664, and there seems to be a lack of review or consensus around most changes. Even in the recent changes, the review brought up several points that were seemingly ignored, and more that are still being ignored. The constant churn and reverts on this API is not acceptable. -- Regards, Bryan Drewery ___ 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: r303111 - head/share/man/man4
Author: mav Date: Wed Jul 20 20:10:37 2016 New Revision: 303111 URL: https://svnweb.freebsd.org/changeset/base/303111 Log: Document list of supported chipsets. Modified: head/share/man/man4/intpm.4 Modified: head/share/man/man4/intpm.4 == --- head/share/man/man4/intpm.4 Wed Jul 20 19:21:11 2016(r303110) +++ head/share/man/man4/intpm.4 Wed Jul 20 20:10:37 2016(r303111) @@ -24,25 +24,39 @@ .\" .\" $FreeBSD$ .\" -.Dd January 8, 1999 +.Dd July 20, 2016 .Dt INTPM 4 .Os .Sh NAME .Nm intpm .Nd Intel PIIX4 Power Management controller driver .Sh SYNOPSIS +.Cd device pci .Cd device smbus .Cd device smb .Cd device intpm .Sh DESCRIPTION -This driver provides access to -.Tn Intel PIIX4 PCI Controller function 3 , -Power management controller. -Currently, only smbus controller -function is implemented. -But it also have bus idle monitoring function. -It -will display mapped I/O address for bus monitoring function when attaching. +The +.Nm +driver provides access to +.Tn Intel PIIX4 +compatible Power Management controllers. +Currently, only +.Xr smbus 4 +controller function is implemented. +.Sh HARDWARE +The +.Nm +driver supports the following chipsets: +.Pp +.Bl -bullet -compact +.It +Intel 82371AB/82443MX +.It +ATI IXP400 +.It +AMD SB600/700/710/750 +.El .Sh SEE ALSO .Xr smb 4 , .Xr smbus 4 ___ 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: r303112 - stable/11/lib/libc/locale
Author: bapt Date: Wed Jul 20 20:12:58 2016 New Revision: 303112 URL: https://svnweb.freebsd.org/changeset/base/303112 Log: MFC: r302916 Revert 302324 and properly fix the crash with ISO-8859-5 locales PR: 211135 Reported by: jkim Tested by:jkim Approved by: re (gjb) Modified: stable/11/lib/libc/locale/collate.c Modified: stable/11/lib/libc/locale/collate.c == --- stable/11/lib/libc/locale/collate.c Wed Jul 20 20:10:37 2016 (r303111) +++ stable/11/lib/libc/locale/collate.c Wed Jul 20 20:12:58 2016 (r303112) @@ -310,7 +310,7 @@ _collate_lookup(struct xlocale_collate * if ((sptr = *state) != NULL) { *pri = *sptr; sptr++; - if ((sptr == *state) || (*sptr == 0)) + if ((sptr == *state) || (sptr == NULL)) *state = NULL; else *state = sptr; @@ -451,6 +451,7 @@ _collate_wxfrm(struct xlocale_collate *t errno = EINVAL; goto fail; } + state = NULL; pri = COLLATE_MAX_PRIORITY; } if (room) { @@ -469,6 +470,7 @@ _collate_wxfrm(struct xlocale_collate *t errno = EINVAL; goto fail; } + state = NULL; continue; } if (room) { @@ -597,6 +599,7 @@ _collate_sxfrm(struct xlocale_collate *t errno = EINVAL; goto fail; } + state = NULL; pri = COLLATE_MAX_PRIORITY; } @@ -622,6 +625,7 @@ _collate_sxfrm(struct xlocale_collate *t errno = EINVAL; goto fail; } + state = NULL; continue; } ___ 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: r303113 - head/share/man/man4
Author: mav Date: Wed Jul 20 20:18:52 2016 New Revision: 303113 URL: https://svnweb.freebsd.org/changeset/base/303113 Log: Cross-link some SMBus controller drivers man pages. Some compatibility cases are not obvious. Modified: head/share/man/man4/amdpm.4 head/share/man/man4/amdsmb.4 head/share/man/man4/ichsmb.4 head/share/man/man4/intpm.4 head/share/man/man4/ismt.4 Modified: head/share/man/man4/amdpm.4 == --- head/share/man/man4/amdpm.4 Wed Jul 20 20:12:58 2016(r303112) +++ head/share/man/man4/amdpm.4 Wed Jul 20 20:18:52 2016(r303113) @@ -51,6 +51,7 @@ for writing user code to fetch voltages, monitoring chip of your mainboard. .Sh SEE ALSO .Xr amdsmb 4 , +.Xr intpm 4 , .Xr smb 4 , .Xr smbus 4 .Sh HISTORY Modified: head/share/man/man4/amdsmb.4 == --- head/share/man/man4/amdsmb.4Wed Jul 20 20:12:58 2016 (r303112) +++ head/share/man/man4/amdsmb.4Wed Jul 20 20:18:52 2016 (r303113) @@ -31,6 +31,7 @@ .Nm amdsmb .Nd "AMD-8111 SMBus 2.0 controller driver" .Sh SYNOPSIS +.Cd "device pci" .Cd "device smbus" .Cd "device smb" .Cd "device amdsmb" @@ -39,6 +40,8 @@ The .Nm driver provides access to the AMD-8111 SMBus 2.0 controller. .Sh SEE ALSO +.Xr amdpm 4 , +.Xr intpm 4 , .Xr smb 4 , .Xr smbus 4 .Sh HISTORY Modified: head/share/man/man4/ichsmb.4 == --- head/share/man/man4/ichsmb.4Wed Jul 20 20:12:58 2016 (r303112) +++ head/share/man/man4/ichsmb.4Wed Jul 20 20:18:52 2016 (r303113) @@ -53,6 +53,8 @@ driver provides support for the SMBus controller logical device contained in all Intel motherboard chipsets starting from 82801AA (ICH). .Sh SEE ALSO +.Xr intpm 4 , +.Xr ismt 4 , .Xr smb 4 , .Xr smbus 4 .Sh AUTHORS Modified: head/share/man/man4/intpm.4 == --- head/share/man/man4/intpm.4 Wed Jul 20 20:12:58 2016(r303112) +++ head/share/man/man4/intpm.4 Wed Jul 20 20:18:52 2016(r303113) @@ -58,6 +58,9 @@ ATI IXP400 AMD SB600/700/710/750 .El .Sh SEE ALSO +.Xr amdpm 4 , +.Xr amdsmb 4 , +.Xr ichsmb 4 , .Xr smb 4 , .Xr smbus 4 .Sh HISTORY Modified: head/share/man/man4/ismt.4 == --- head/share/man/man4/ismt.4 Wed Jul 20 20:12:58 2016(r303112) +++ head/share/man/man4/ismt.4 Wed Jul 20 20:18:52 2016(r303113) @@ -48,6 +48,7 @@ This driver provides access to the SMBus 2.0 controller device contained in the Intel Atom S1200 and C2000 CPUs. .Sh SEE ALSO +.Xr ichsmb 4 , .Xr smb 4 , .Xr smbus 4 .Sh HISTORY ___ 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: r303114 - head/usr.bin/gcore
Author: cem Date: Wed Jul 20 22:03:45 2016 New Revision: 303114 URL: https://svnweb.freebsd.org/changeset/base/303114 Log: Add extended segment support to gcore A follow-up to r303099, D7255. Basically, apply the exact same change, with the exact same rationale, to gcore. gcore's elfcore.c is largely a clone of the kernel imgact_elf coredump facility. Reviewed by: emaste (earlier version, not substantially different) Requested by: jhb Sponsored by: EMC / Isilon Storage Division Differential Revision:https://reviews.freebsd.org/D7265 Modified: head/usr.bin/gcore/elfcore.c Modified: head/usr.bin/gcore/elfcore.c == --- head/usr.bin/gcore/elfcore.cWed Jul 20 20:18:52 2016 (r303113) +++ head/usr.bin/gcore/elfcore.cWed Jul 20 22:03:45 2016 (r303114) @@ -213,6 +213,8 @@ elf_coredump(int efd __unused, int fd, p */ sb = sbuf_new_auto(); hdrsize = sizeof(Elf_Ehdr) + sizeof(Elf_Phdr) * (1 + seginfo.count); + if (seginfo.count + 1 >= PN_XNUM) + hdrsize += sizeof(Elf_Shdr); /* Start header + notes section. */ sbuf_start_section(sb, NULL); /* Make empty header subsection. */ @@ -423,10 +425,10 @@ elf_puthdr(pid_t pid, vm_map_entry_t map { Elf_Ehdr *ehdr; Elf_Phdr *phdr; + Elf_Shdr *shdr; struct phdr_closure phc; ehdr = (Elf_Ehdr *)hdr; - phdr = (Elf_Phdr *)((char *)hdr + sizeof(Elf_Ehdr)); ehdr->e_ident[EI_MAG0] = ELFMAG0; ehdr->e_ident[EI_MAG1] = ELFMAG1; @@ -446,14 +448,40 @@ elf_puthdr(pid_t pid, vm_map_entry_t map ehdr->e_flags = 0; ehdr->e_ehsize = sizeof(Elf_Ehdr); ehdr->e_phentsize = sizeof(Elf_Phdr); - ehdr->e_phnum = numsegs + 1; ehdr->e_shentsize = sizeof(Elf_Shdr); - ehdr->e_shnum = 0; ehdr->e_shstrndx = SHN_UNDEF; + if (numsegs + 1 < PN_XNUM) { + ehdr->e_phnum = numsegs + 1; + ehdr->e_shnum = 0; + } else { + ehdr->e_phnum = PN_XNUM; + ehdr->e_shnum = 1; + + ehdr->e_shoff = ehdr->e_phoff + + (numsegs + 1) * ehdr->e_phentsize; + + shdr = (Elf_Shdr *)((char *)hdr + ehdr->e_shoff); + memset(shdr, 0, sizeof(*shdr)); + /* +* A special first section is used to hold large segment and +* section counts. This was proposed by Sun Microsystems in +* Solaris and has been adopted by Linux; the standard ELF +* tools are already familiar with the technique. +* +* See table 7-7 of the Solaris "Linker and Libraries Guide" +* (or 12-7 depending on the version of the document) for more +* details. +*/ + shdr->sh_type = SHT_NULL; + shdr->sh_size = ehdr->e_shnum; + shdr->sh_link = ehdr->e_shstrndx; + shdr->sh_info = numsegs + 1; + } /* * Fill in the program header entries. */ + phdr = (Elf_Phdr *)((char *)hdr + ehdr->e_phoff); /* The note segement. */ phdr->p_type = PT_NOTE; ___ 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: r303115 - head/sys/dev/urtwn
Author: avos Date: Wed Jul 20 22:05:44 2016 New Revision: 303115 URL: https://svnweb.freebsd.org/changeset/base/303115 Log: urtwn: fix A-MSDU tx aggregation when encryption is used. Clear unused (undocumented) CAM bytes while setting a key; without that, hardware does weird things when A-MSDU bit in QoS header is set. Tested with RTL8188CUS (AP) -> RTL8188EU (STA) (A-MSDU transmit). Reported by: many Obtained from:https://github.com/s3erios/urtwm MFC after:5 days Modified: head/sys/dev/urtwn/if_urtwn.c head/sys/dev/urtwn/if_urtwnreg.h Modified: head/sys/dev/urtwn/if_urtwn.c == --- head/sys/dev/urtwn/if_urtwn.c Wed Jul 20 22:03:45 2016 (r303114) +++ head/sys/dev/urtwn/if_urtwn.c Wed Jul 20 22:05:44 2016 (r303115) @@ -2318,6 +2318,10 @@ urtwn_key_set_cb(struct urtwn_softc *sc, k->wk_cipher->ic_cipher, algo, k->wk_flags, k->wk_keylen, ether_sprintf(k->wk_macaddr)); + /* Clear high bits. */ + urtwn_cam_write(sc, R92C_CAM_CTL6(k->wk_keyix), 0); + urtwn_cam_write(sc, R92C_CAM_CTL7(k->wk_keyix), 0); + /* Write key. */ for (i = 0; i < 4; i++) { error = urtwn_cam_write(sc, R92C_CAM_KEY(k->wk_keyix, i), Modified: head/sys/dev/urtwn/if_urtwnreg.h == --- head/sys/dev/urtwn/if_urtwnreg.hWed Jul 20 22:03:45 2016 (r303114) +++ head/sys/dev/urtwn/if_urtwnreg.hWed Jul 20 22:05:44 2016 (r303115) @@ -871,6 +871,8 @@ #define R92C_CAM_CTL0(entry) ((entry) * 8 + 0) #define R92C_CAM_CTL1(entry) ((entry) * 8 + 1) #define R92C_CAM_KEY(entry, i) ((entry) * 8 + 2 + (i)) +#define R92C_CAM_CTL6(entry) ((entry) * 8 + 6) +#define R92C_CAM_CTL7(entry) ((entry) * 8 + 7) /* Bits for R92C_CAM_CTL0(i). */ #define R92C_CAM_KEYID_M 0x0003 ___ 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: r303116 - stable/10/sys/dev/e1000
Author: sbruno Date: Wed Jul 20 22:28:26 2016 New Revision: 303116 URL: https://svnweb.freebsd.org/changeset/base/303116 Log: Fix em(4) build after r303107 Reported by: Jason Wolf Modified: stable/10/sys/dev/e1000/if_em.c Modified: stable/10/sys/dev/e1000/if_em.c == --- stable/10/sys/dev/e1000/if_em.c Wed Jul 20 22:05:44 2016 (r303115) +++ stable/10/sys/dev/e1000/if_em.c Wed Jul 20 22:28:26 2016 (r303116) @@ -1220,7 +1220,7 @@ em_ioctl(struct ifnet *ifp, u_long comma ifp->if_mtu = ifr->ifr_mtu; adapter->hw.mac.max_frame_size = ifp->if_mtu + ETHER_HDR_LEN + ETHER_CRC_LEN; - if (if_getdrvflags(ifp) & IFF_DRV_RUNNING) + if (ifp->if_drv_flags & IFF_DRV_RUNNING) em_init_locked(adapter); EM_CORE_UNLOCK(adapter); break; ___ 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: r303117 - stable/10/sys/dev/e1000
Author: sbruno Date: Wed Jul 20 22:32:19 2016 New Revision: 303117 URL: https://svnweb.freebsd.org/changeset/base/303117 Log: Fix lem(4) build after r303107 Submitted by: Jason Wolfe Modified: stable/10/sys/dev/e1000/if_lem.c Modified: stable/10/sys/dev/e1000/if_lem.c == --- stable/10/sys/dev/e1000/if_lem.cWed Jul 20 22:28:26 2016 (r303116) +++ stable/10/sys/dev/e1000/if_lem.cWed Jul 20 22:32:19 2016 (r303117) @@ -1047,7 +1047,7 @@ lem_ioctl(struct ifnet *ifp, u_long comm ifp->if_mtu = ifr->ifr_mtu; adapter->max_frame_size = ifp->if_mtu + ETHER_HDR_LEN + ETHER_CRC_LEN; - if ((if_getdrvflags(ifp) & IFF_DRV_RUNNING)) + if (ifp->if_drv_flags & IFF_DRV_RUNNING) lem_init_locked(adapter); EM_CORE_UNLOCK(adapter); break; ___ 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: r303118 - head/sys/kern
Author: cem Date: Wed Jul 20 22:46:56 2016 New Revision: 303118 URL: https://svnweb.freebsd.org/changeset/base/303118 Log: ANSI-fy imgact_elf.c Sponsored by: EMC / Isilon Storage Division Modified: head/sys/kern/imgact_elf.c Modified: head/sys/kern/imgact_elf.c == --- head/sys/kern/imgact_elf.c Wed Jul 20 22:32:19 2016(r303117) +++ head/sys/kern/imgact_elf.c Wed Jul 20 22:46:56 2016(r303118) @@ -1449,9 +1449,7 @@ cb_put_phdr(entry, closure) * the number of segments and their total size. */ static void -cb_size_segment(entry, closure) - vm_map_entry_t entry; - void *closure; +cb_size_segment(vm_map_entry_t entry, void *closure) { struct sseg_closure *ssc = (struct sseg_closure *)closure; @@ -1465,10 +1463,7 @@ cb_size_segment(entry, closure) * caller-supplied data. */ static void -each_writable_segment(td, func, closure) - struct thread *td; - segment_callback func; - void *closure; +each_writable_segment(struct thread *td, segment_callback func, void *closure) { struct proc *p = td->td_proc; vm_map_t map = &p->p_vmspace->vm_map; ___ 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: r303119 - head/release/tools
Author: gjb Date: Wed Jul 20 22:49:33 2016 New Revision: 303119 URL: https://svnweb.freebsd.org/changeset/base/303119 Log: Create a /usr/home -> /home symlink for the arm images to avoid /usr/home confusingly being created as a directory. Reported by: Russell Haley MFC after:3 days Sponsored by: The FreeBSD Foundation Modified: head/release/tools/arm.subr Modified: head/release/tools/arm.subr == --- head/release/tools/arm.subr Wed Jul 20 22:46:56 2016(r303118) +++ head/release/tools/arm.subr Wed Jul 20 22:49:33 2016(r303119) @@ -88,6 +88,7 @@ arm_create_user() { -c 'FreeBSD User' -d '/home/freebsd' -s '/bin/csh' chroot ${CHROOTDIR} /usr/sbin/pw -R ${DESTDIR} \ usermod root -w yes + chroot ${CHROOTDIR} ln -s /home ${DESTDIR}/usr/home return 0 } ___ 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: r303120 - head/sys/kern
Author: cem Date: Wed Jul 20 22:51:33 2016 New Revision: 303120 URL: https://svnweb.freebsd.org/changeset/base/303120 Log: imgact_elf: Rename the segment iterator to match reality The each_writable_segment routine evaluates segments on a slightly little more nuanced metric than simply "writable" or not. Rename the function to more closely match its behavior (each_dumpable_segment). Suggested by: jhb Sponsored by: EMC / Isilon Storage Division Modified: head/sys/kern/imgact_elf.c Modified: head/sys/kern/imgact_elf.c == --- head/sys/kern/imgact_elf.c Wed Jul 20 22:49:33 2016(r303119) +++ head/sys/kern/imgact_elf.c Wed Jul 20 22:51:33 2016(r303120) @@ -1163,7 +1163,7 @@ static void cb_put_phdr(vm_map_entry_t, static void cb_size_segment(vm_map_entry_t, void *); static int core_write(struct coredump_params *, void *, size_t, off_t, enum uio_seg); -static void each_writable_segment(struct thread *, segment_callback, void *); +static void each_dumpable_segment(struct thread *, segment_callback, void *); static int __elfN(corehdr)(struct coredump_params *, int, void *, size_t, struct note_info_list *, size_t); static void __elfN(prepare_notes)(struct thread *, struct note_info_list *, @@ -1317,7 +1317,7 @@ __elfN(coredump)(struct thread *td, stru /* Size the program segments. */ seginfo.count = 0; seginfo.size = 0; - each_writable_segment(td, cb_size_segment, &seginfo); + each_dumpable_segment(td, cb_size_segment, &seginfo); /* * Collect info about the core file header area. @@ -1419,7 +1419,7 @@ done: } /* - * A callback for each_writable_segment() to write out the segment's + * A callback for each_dumpable_segment() to write out the segment's * program header entry. */ static void @@ -1445,7 +1445,7 @@ cb_put_phdr(entry, closure) } /* - * A callback for each_writable_segment() to gather information about + * A callback for each_dumpable_segment() to gather information about * the number of segments and their total size. */ static void @@ -1463,7 +1463,7 @@ cb_size_segment(vm_map_entry_t entry, vo * caller-supplied data. */ static void -each_writable_segment(struct thread *td, segment_callback func, void *closure) +each_dumpable_segment(struct thread *td, segment_callback func, void *closure) { struct proc *p = td->td_proc; vm_map_t map = &p->p_vmspace->vm_map; @@ -1694,7 +1694,7 @@ __elfN(puthdr)(struct thread *td, void * /* All the writable segments from the program. */ phc.phdr = phdr; phc.offset = round_page(hdrsize + notesz); - each_writable_segment(td, cb_put_phdr, &phc); + each_dumpable_segment(td, cb_put_phdr, &phc); } static size_t ___ 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: r303120 - head/sys/kern
On Wednesday, July 20, 2016 10:51:33 PM Conrad E. Meyer wrote: > Author: cem > Date: Wed Jul 20 22:51:33 2016 > New Revision: 303120 > URL: https://svnweb.freebsd.org/changeset/base/303120 > > Log: > imgact_elf: Rename the segment iterator to match reality > > The each_writable_segment routine evaluates segments on a slightly little > more > nuanced metric than simply "writable" or not. Rename the function to more > closely match its behavior (each_dumpable_segment). > > Suggested by: jhb > Sponsored by: EMC / Isilon Storage Division Please do the same rename in gcore's elfcore.c as well. -- 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: r303043 - in head: share/man/man4 sys/dev/vt sys/dev/vt/hw/fb
On 21/07/2016 5:07 AM, Bryan Drewery wrote: > On 7/20/16 8:20 AM, Conrad Meyer wrote: >> On Wed, Jul 20, 2016 at 12:44 AM, Kubilay Kocak wrote: >>> On 20/07/2016 5:20 AM, Conrad E. Meyer wrote: Author: cem Date: Tue Jul 19 19:20:47 2016 New Revision: 303043 URL: https://svnweb.freebsd.org/changeset/base/303043 Log: Increase vt(4) framebuffer maximum size And rename "DEFAULT" constants to the more accurate "MAX." PR: 210382 Submitted by: Felix Reviewed by:wblock, cem Tested by: Dave Cottlehuber >>> >>> Can MFC? >>> >>> There are high hopes for this coming out in 11.0 / next 10.x >>> >>> re is cc'd on the original bugzilla issue >> >> If you would like to do so, please go ahead. >> >> Best, >> Conrad >> > > I'll handle it after it's been in for a week. > Thank you Bryan ./koobs ___ 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: r303121 - stable/10/release
Author: gjb Date: Thu Jul 21 00:52:53 2016 New Revision: 303121 URL: https://svnweb.freebsd.org/changeset/base/303121 Log: MFC r303003: Make sure we clean up the rdoc and reldoc directories if NODOC is not set. Sponsored by: The FreeBSD Foundation Modified: stable/10/release/Makefile Directory Properties: stable/10/ (props changed) Modified: stable/10/release/Makefile == --- stable/10/release/Makefile Wed Jul 20 22:51:33 2016(r303120) +++ stable/10/release/Makefile Thu Jul 21 00:52:53 2016(r303121) @@ -137,6 +137,9 @@ CLEANFILES+=${I}.xz CLEANFILES+= pkg-stage .endif CLEANDIRS= dist ftp disc1 bootonly dvd +.if !defined(NODOC) +CLEANDIRS+=reldoc rdoc +.endif beforeclean: chflags -R noschg . .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: r303122 - head/usr.bin/gcore
Author: cem Date: Thu Jul 21 00:53:14 2016 New Revision: 303122 URL: https://svnweb.freebsd.org/changeset/base/303122 Log: Mirror r303120 to gcore Sponsored by: EMC / Isilon Storage Division Modified: head/usr.bin/gcore/elfcore.c Modified: head/usr.bin/gcore/elfcore.c == --- head/usr.bin/gcore/elfcore.cThu Jul 21 00:52:53 2016 (r303121) +++ head/usr.bin/gcore/elfcore.cThu Jul 21 00:53:14 2016 (r303122) @@ -95,7 +95,7 @@ typedef void* (*notefunc_t)(void *, size static void cb_put_phdr(vm_map_entry_t, void *); static void cb_size_segment(vm_map_entry_t, void *); -static void each_writable_segment(vm_map_entry_t, segment_callback, +static void each_dumpable_segment(vm_map_entry_t, segment_callback, void *closure); static void elf_detach(void); /* atexit() handler. */ static void *elf_note_fpregset(void *, size_t *); @@ -206,7 +206,7 @@ elf_coredump(int efd __unused, int fd, p /* Size the program segments. */ seginfo.count = 0; seginfo.size = 0; - each_writable_segment(map, cb_size_segment, &seginfo); + each_dumpable_segment(map, cb_size_segment, &seginfo); /* * Build the header and the notes using sbuf and write to the file. @@ -277,7 +277,7 @@ elf_coredump(int efd __unused, int fd, p } /* - * A callback for each_writable_segment() to write out the segment's + * A callback for each_dumpable_segment() to write out the segment's * program header entry. */ static void @@ -307,7 +307,7 @@ cb_put_phdr(vm_map_entry_t entry, void * } /* - * A callback for each_writable_segment() to gather information about + * A callback for each_dumpable_segment() to gather information about * the number of segments and their total size. */ static void @@ -325,7 +325,7 @@ cb_size_segment(vm_map_entry_t entry, vo * data. */ static void -each_writable_segment(vm_map_entry_t map, segment_callback func, void *closure) +each_dumpable_segment(vm_map_entry_t map, segment_callback func, void *closure) { vm_map_entry_t entry; @@ -497,7 +497,7 @@ elf_puthdr(pid_t pid, vm_map_entry_t map /* All the writable segments from the program. */ phc.phdr = phdr; phc.offset = segoff; - each_writable_segment(map, cb_put_phdr, &phc); + each_dumpable_segment(map, cb_put_phdr, &phc); } /* ___ 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: r303120 - head/sys/kern
Sure — r303122. Conrad On Wed, Jul 20, 2016 at 4:09 PM, John Baldwin wrote: > On Wednesday, July 20, 2016 10:51:33 PM Conrad E. Meyer wrote: >> Author: cem >> Date: Wed Jul 20 22:51:33 2016 >> New Revision: 303120 >> URL: https://svnweb.freebsd.org/changeset/base/303120 >> >> Log: >> imgact_elf: Rename the segment iterator to match reality >> >> The each_writable_segment routine evaluates segments on a slightly little >> more >> nuanced metric than simply "writable" or not. Rename the function to more >> closely match its behavior (each_dumpable_segment). >> >> Suggested by: jhb >> Sponsored by: EMC / Isilon Storage Division > > Please do the same rename in gcore's elfcore.c as well. > > -- > 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"
svn commit: r303123 - head/sys/cam
Author: imp Date: Thu Jul 21 03:11:35 2016 New Revision: 303123 URL: https://svnweb.freebsd.org/changeset/base/303123 Log: Fix mismerge and include the nvme support. Also, print out the name of any CCB's functions that's not supported. MFC after: 1 week Modified: head/sys/cam/cam_xpt.c Modified: head/sys/cam/cam_xpt.c == --- head/sys/cam/cam_xpt.c Thu Jul 21 00:53:14 2016(r303122) +++ head/sys/cam/cam_xpt.c Thu Jul 21 03:11:35 2016(r303123) @@ -1033,6 +1033,8 @@ xpt_announce_periph(struct cam_periph *p else if (path->device->protocol == PROTO_SEMB) semb_print_ident( (struct sep_identify_data *)&path->device->ident_data); + else if (path->device->protocol == PROTO_NVME) + nvme_print_ident(path->device->nvme_cdata, path->device->nvme_data); else printf("Unknown protocol device\n"); if (path->device->serial_num_len > 0) { @@ -1086,6 +1088,8 @@ xpt_denounce_periph(struct cam_periph *p else if (path->device->protocol == PROTO_SEMB) semb_print_ident_short( (struct sep_identify_data *)&path->device->ident_data); + else if (path->device->protocol == PROTO_NVME) + nvme_print_ident(path->device->nvme_cdata, path->device->nvme_data); else printf("Unknown protocol device"); if (path->device->serial_num_len > 0) @@ -2516,6 +2520,10 @@ xpt_action_default(union ccb *start_ccb) if (start_ccb->ccb_h.func_code == XPT_ATA_IO) start_ccb->ataio.resid = 0; /* FALLTHROUGH */ + case XPT_NVME_IO: + if (start_ccb->ccb_h.func_code == XPT_NVME_IO) + start_ccb->nvmeio.resid = 0; + /* FALLTHROUGH */ case XPT_RESET_DEV: case XPT_ENG_EXEC: case XPT_SMP_IO: @@ -2655,6 +2663,8 @@ call_sim: cgd->inq_data = dev->inq_data; cgd->ident_data = dev->ident_data; cgd->inq_flags = dev->inq_flags; + cgd->nvme_data = dev->nvme_data; + cgd->nvme_cdata = dev->nvme_cdata; cgd->ccb_h.status = CAM_REQ_CMP; cgd->serial_num_len = dev->serial_num_len; if ((dev->serial_num_len > 0) @@ -3011,8 +3021,10 @@ call_sim: case XPT_TERM_IO: case XPT_ENG_INQ: /* XXX Implement */ - printf("%s: CCB type %#x not supported\n", __func__, - start_ccb->ccb_h.func_code); + xpt_print_path(start_ccb->ccb_h.path); + printf("%s: CCB type %#x %s not supported\n", __func__, + start_ccb->ccb_h.func_code, + xpt_action_name(start_ccb->ccb_h.func_code)); start_ccb->ccb_h.status = CAM_PROVIDE_FAIL; if (start_ccb->ccb_h.func_code & XPT_FC_DEV_QUEUED) { xpt_done(start_ccb); @@ -3315,6 +3327,13 @@ xpt_run_devq(struct cam_devq *devq) ata_cmd_string(&work_ccb->ataio.cmd, cdb_str, sizeof(cdb_str; break; + case XPT_NVME_IO: + CAM_DEBUG(work_ccb->ccb_h.path, + CAM_DEBUG_CDB,("%s. NCB: %s\n", +nvme_op_string(&work_ccb->nvmeio.cmd), +nvme_cmd_string(&work_ccb->nvmeio.cmd, + cdb_str, sizeof(cdb_str; + break; default: break; } @@ -3904,6 +3923,9 @@ xpt_bus_register(struct cam_sim *sim, de case XPORT_SATA: new_bus->xport = ata_get_xport(); break; + case XPORT_NVME: + new_bus->xport = nvme_get_xport(); + break; default: new_bus->xport = &xport_default; break; ___ 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: r303126 - head/sys/dev/nvme
Author: imp Date: Thu Jul 21 03:11:39 2016 New Revision: 303126 URL: https://svnweb.freebsd.org/changeset/base/303126 Log: Actually import nvme_sim so the CAM attachment for NVME (nda) actually works. MFC after: 1 week Added: head/sys/dev/nvme/nvme_sim.c (contents, props changed) Modified: head/sys/dev/nvme/nvme_private.h Modified: head/sys/dev/nvme/nvme_private.h == --- head/sys/dev/nvme/nvme_private.hThu Jul 21 03:11:37 2016 (r303125) +++ head/sys/dev/nvme/nvme_private.hThu Jul 21 03:11:39 2016 (r303126) @@ -245,10 +245,6 @@ struct nvme_controller { struct mtx lock; - struct cam_sim *sim; - struct cam_path *path; - int cam_ref; - uint32_tready_timeout_in_ms; bus_space_tag_t bus_tag; Added: head/sys/dev/nvme/nvme_sim.c == --- /dev/null 00:00:00 1970 (empty, because file is newly added) +++ head/sys/dev/nvme/nvme_sim.cThu Jul 21 03:11:39 2016 (r303126) @@ -0,0 +1,400 @@ +/*- + * Copyright (c) 2016 Netflix, 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, + *without modification, immediately at the beginning of the file. + * 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 +#include +#include +#include + +#include +#include +#include +#include +#include // Yes, this is wrong. +#include + +#include "nvme_private.h" + +#define ccb_accb_ptr spriv_ptr0 +#define ccb_ctrlr_ptr spriv_ptr1 +static voidnvme_sim_action(struct cam_sim *sim, union ccb *ccb); +static voidnvme_sim_poll(struct cam_sim *sim); + +#define sim2softc(sim) ((struct nvme_sim_softc *)cam_sim_softc(sim)) +#define sim2ns(sim)(sim2softc(sim)->s_ns) +#define sim2ctrlr(sim) (sim2softc(sim)->s_ctrlr) + +struct nvme_sim_softc +{ + struct nvme_controller *s_ctrlr; + struct nvme_namespace *s_ns; + struct cam_sim *s_sim; + struct cam_path *s_path; +}; + +static void +nvme_sim_nvmeio_done(void *ccb_arg, const struct nvme_completion *cpl) +{ + union ccb *ccb = (union ccb *)ccb_arg; + + /* +* Let the periph know the completion, and let it sort out what +* it means. Make our best guess, though for the status code. +*/ + memcpy(&ccb->nvmeio.cpl, cpl, sizeof(*cpl)); + if (nvme_completion_is_error(cpl)) + ccb->ccb_h.status = CAM_REQ_CMP_ERR; + else + ccb->ccb_h.status = CAM_REQ_CMP; + xpt_done(ccb); +} + +static void +nvme_sim_nvmeio(struct cam_sim *sim, union ccb *ccb) +{ + struct ccb_nvmeio *nvmeio = &ccb->nvmeio; + struct nvme_request *req; + void*payload; + uint32_tsize; + struct nvme_controller *ctrlr; + + ctrlr = sim2ctrlr(sim); + payload = nvmeio->data_ptr; + size = nvmeio->dxfer_len; + /* SG LIST ??? */ + if ((nvmeio->ccb_h.flags & CAM_DATA_MASK) == CAM_DATA_BIO) + req = nvme_allocate_request_bio((struct bio *)payload, + nvme_sim_nvmeio_done, ccb); + else if (payload == NULL) + req = nvme_allocate_request_null(nvme_sim_nvmeio_done, ccb); + else + req = nvme_allocate_request_vaddr(payload, size, + nvme_sim_nvmeio_done, ccb); + + if (req == NULL) { + nvmeio->ccb_h.status = CAM_RESRC_UNAVAIL; + xpt_done(ccb); + return; + } + + memcpy(&r
svn commit: r303125 - head/sys/cam/nvme
Author: imp Date: Thu Jul 21 03:11:37 2016 New Revision: 303125 URL: https://svnweb.freebsd.org/changeset/base/303125 Log: Remove some bogus comments and printfs. Also, we can't cam_periph_releaes_locked() at the end of nvme_probe_start because we hit an assertion which may be bogus. Instead, leak a periph until we sort it out. Since these devices don't arrive and depart often, so this is the lessor of two evils. MFC after: 1 week Modified: head/sys/cam/nvme/nvme_xpt.c Modified: head/sys/cam/nvme/nvme_xpt.c == --- head/sys/cam/nvme/nvme_xpt.cThu Jul 21 03:11:36 2016 (r303124) +++ head/sys/cam/nvme/nvme_xpt.cThu Jul 21 03:11:37 2016 (r303125) @@ -163,13 +163,14 @@ static struct xpt_xport nvme_xport = { struct xpt_xport * nvme_get_xport(void) { + return (&nvme_xport); } static void nvme_probe_periph_init() { - printf("nvme cam probe device init\n"); + } static cam_status @@ -297,18 +298,15 @@ nvme_probe_start(struct cam_periph *peri start_ccb->ccb_h.status = CAM_REQ_CMP; xpt_done(start_ccb); } -// XXX not sure I need this -// XXX unlike other XPTs, we never freeze the queue since we have a super-simple -// XXX state machine - /* Drop freeze taken due to CAM_DEV_QFREEZE flag set. -- did we really do this? */ -// cam_release_devq(path, 0, 0, 0, FALSE); cam_periph_invalidate(periph); - cam_periph_release_locked(periph); + /* Can't release periph since we hit a (possibly bogus) assertion */ +// cam_periph_release_locked(periph); } static void nvme_probe_cleanup(struct cam_periph *periph) { + free(periph->softc, M_CAMXPT); } ___ 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: r303124 - head/sys/cam
Author: imp Date: Thu Jul 21 03:11:36 2016 New Revision: 303124 URL: https://svnweb.freebsd.org/changeset/base/303124 Log: Fix mismerge and add NVME xport support to devstat (or rather, exclude it from SCSI). MFC after: 1 week Modified: head/sys/cam/cam_ccb.h Modified: head/sys/cam/cam_ccb.h == --- head/sys/cam/cam_ccb.h Thu Jul 21 03:11:35 2016(r303123) +++ head/sys/cam/cam_ccb.h Thu Jul 21 03:11:36 2016(r303124) @@ -285,10 +285,11 @@ typedef enum { XPORT_NVME, /* NVMe over PCIe */ } cam_xport; +#define XPORT_IS_NVME(t) ((t) == XPORT_NVME) #define XPORT_IS_ATA(t)((t) == XPORT_ATA || (t) == XPORT_SATA) #define XPORT_IS_SCSI(t) ((t) != XPORT_UNKNOWN && \ (t) != XPORT_UNSPECIFIED && \ -!XPORT_IS_ATA(t)) +!XPORT_IS_ATA(t) && !XPORT_IS_NVME(t)) #define XPORT_DEVSTAT_TYPE(t) (XPORT_IS_ATA(t) ? DEVSTAT_TYPE_IF_IDE : \ XPORT_IS_SCSI(t) ? DEVSTAT_TYPE_IF_SCSI : \ DEVSTAT_TYPE_IF_OTHER) ___ 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: r303127 - head/sys/dev/hyperv/vmbus
Author: sephe Date: Thu Jul 21 05:30:31 2016 New Revision: 303127 URL: https://svnweb.freebsd.org/changeset/base/303127 Log: hyperv/vmbus: Save event flag location and evet flag mask. This avoids unnecessary access to the vmbus_softc struct on sending path. MFC after:1 week Sponsored by: Microsoft OSTC Differential Revision:https://reviews.freebsd.org/D7257 Modified: head/sys/dev/hyperv/vmbus/vmbus_chan.c head/sys/dev/hyperv/vmbus/vmbus_chanvar.h Modified: head/sys/dev/hyperv/vmbus/vmbus_chan.c == --- head/sys/dev/hyperv/vmbus/vmbus_chan.c Thu Jul 21 03:11:39 2016 (r303126) +++ head/sys/dev/hyperv/vmbus/vmbus_chan.c Thu Jul 21 05:30:31 2016 (r303127) @@ -51,7 +51,6 @@ __FBSDID("$FreeBSD$"); #include #include -static voidvmbus_chan_signal_tx(struct hv_vmbus_channel *chan); static voidvmbus_chan_update_evtflagcnt(struct vmbus_softc *, const struct hv_vmbus_channel *); @@ -77,18 +76,13 @@ vmbus_chan_msgprocs[VMBUS_CHANMSG_TYPE_M VMBUS_CHANMSG_PROC_WAKEUP(GPADL_DISCONNRESP) }; -/** - * @brief Trigger an event notification on the specified channel +/* + * Notify host that there are data pending on our TX bufring. */ -static void -vmbus_chan_signal_tx(struct hv_vmbus_channel *chan) +static __inline void +vmbus_chan_signal_tx(const struct hv_vmbus_channel *chan) { - struct vmbus_softc *sc = chan->ch_vmbus; - uint32_t chanid = chan->ch_id; - - atomic_set_long(&sc->vmbus_tx_evtflags[chanid >> VMBUS_EVTFLAG_SHIFT], - 1UL << (chanid & VMBUS_EVTFLAG_MASK)); - + atomic_set_long(chan->ch_evtflag, chan->ch_evtflag_mask); if (chan->ch_flags & VMBUS_CHAN_FLAG_HASMNF) atomic_set_int(chan->ch_montrig, chan->ch_montrig_mask); else @@ -1120,6 +1114,13 @@ vmbus_chan_msgproc_choffer(struct vmbus_ 1 << (offer->chm_montrig % VMBUS_MONTRIG_LEN); } + /* +* Setup event flag. +*/ + chan->ch_evtflag = + &sc->vmbus_tx_evtflags[chan->ch_id >> VMBUS_EVTFLAG_SHIFT]; + chan->ch_evtflag_mask = 1UL << (chan->ch_id & VMBUS_EVTFLAG_MASK); + /* Select default cpu for this channel. */ vmbus_chan_cpu_default(chan); Modified: head/sys/dev/hyperv/vmbus/vmbus_chanvar.h == --- head/sys/dev/hyperv/vmbus/vmbus_chanvar.h Thu Jul 21 03:11:39 2016 (r303126) +++ head/sys/dev/hyperv/vmbus/vmbus_chanvar.h Thu Jul 21 05:30:31 2016 (r303127) @@ -77,13 +77,20 @@ typedef struct hv_vmbus_channel { uint32_tch_id; /* channel id */ /* -* These are based on the offer_msg.monitor_id. +* These are based on the vmbus_chanmsg_choffer.chm_montrig. * Save it here for easy access. */ - volatile uint32_t *ch_montrig;/* MNF trigger */ + volatile uint32_t *ch_montrig;/* MNF trigger loc. */ uint32_tch_montrig_mask;/* MNF trig mask */ /* +* These are based on the vmbus_chanmsg_choffer.chm_chanid. +* Save it here for easy access. +*/ + volatile u_long *ch_evtflag;/* event flag loc. */ + u_long ch_evtflag_mask;/* event flag */ + + /* * TX bufring; at the beginning of ch_bufring. */ hv_vmbus_ring_buffer_info ch_txbr; ___ 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: r303128 - head/sys/dev/hyperv/vmbus
Author: sephe Date: Thu Jul 21 05:38:05 2016 New Revision: 303128 URL: https://svnweb.freebsd.org/changeset/base/303128 Log: hyperv/vmbus: Reorder channel fields. MFC after:1 week Sponsored by: Microsoft OSTC Differential Revision:https://reviews.freebsd.org/D7258 Modified: head/sys/dev/hyperv/vmbus/vmbus_chan.c head/sys/dev/hyperv/vmbus/vmbus_chanvar.h Modified: head/sys/dev/hyperv/vmbus/vmbus_chan.c == --- head/sys/dev/hyperv/vmbus/vmbus_chan.c Thu Jul 21 05:30:31 2016 (r303127) +++ head/sys/dev/hyperv/vmbus/vmbus_chan.c Thu Jul 21 05:38:05 2016 (r303128) @@ -83,7 +83,7 @@ static __inline void vmbus_chan_signal_tx(const struct hv_vmbus_channel *chan) { atomic_set_long(chan->ch_evtflag, chan->ch_evtflag_mask); - if (chan->ch_flags & VMBUS_CHAN_FLAG_HASMNF) + if (chan->ch_txflags & VMBUS_CHAN_TXF_HASMNF) atomic_set_int(chan->ch_montrig, chan->ch_montrig_mask); else hypercall_signal_event(chan->ch_monprm_dma.hv_paddr); @@ -95,7 +95,7 @@ vmbus_chan_sysctl_mnf(SYSCTL_HANDLER_ARG struct hv_vmbus_channel *chan = arg1; int mnf = 0; - if (chan->ch_flags & VMBUS_CHAN_FLAG_HASMNF) + if (chan->ch_txflags & VMBUS_CHAN_TXF_HASMNF) mnf = 1; return sysctl_handle_int(oidp, &mnf, 0, req); } @@ -1102,7 +1102,7 @@ vmbus_chan_msgproc_choffer(struct vmbus_ /* * Setup MNF stuffs. */ - chan->ch_flags |= VMBUS_CHAN_FLAG_HASMNF; + chan->ch_txflags |= VMBUS_CHAN_TXF_HASMNF; trig_idx = offer->chm_montrig / VMBUS_MONTRIG_LEN; if (trig_idx >= VMBUS_MONTRIGS_MAX) Modified: head/sys/dev/hyperv/vmbus/vmbus_chanvar.h == --- head/sys/dev/hyperv/vmbus/vmbus_chanvar.h Thu Jul 21 05:30:31 2016 (r303127) +++ head/sys/dev/hyperv/vmbus/vmbus_chanvar.h Thu Jul 21 05:38:05 2016 (r303128) @@ -71,42 +71,63 @@ typedef struct { } hv_vmbus_ring_buffer_info; typedef struct hv_vmbus_channel { - device_tch_dev; - struct vmbus_softc *ch_vmbus; + /* +* NOTE: +* Fields before ch_txbr are only accessed on this channel's +* target CPU. +*/ uint32_tch_flags; /* VMBUS_CHAN_FLAG_ */ - uint32_tch_id; /* channel id */ + + /* +* RX bufring; immediately following ch_txbr. +*/ + hv_vmbus_ring_buffer_info ch_rxbr; + + struct taskqueue*ch_tq; + struct task ch_task; + vmbus_chan_callback_t ch_cb; + void*ch_cbarg; + + /* +* TX bufring; at the beginning of ch_bufring. +* +* NOTE: +* Put TX bufring and the following MNF/evtflag to a new +* cacheline, since they will be accessed on all CPUs by +* locking ch_txbr first. +* +* XXX +* TX bufring and following MNF/evtflags do _not_ fit in +* one 64B cacheline. +*/ + hv_vmbus_ring_buffer_info ch_txbr __aligned(CACHE_LINE_SIZE); + uint32_tch_txflags; /* VMBUS_CHAN_TXF_ */ /* * These are based on the vmbus_chanmsg_choffer.chm_montrig. * Save it here for easy access. */ - volatile uint32_t *ch_montrig;/* MNF trigger loc. */ uint32_tch_montrig_mask;/* MNF trig mask */ + volatile uint32_t *ch_montrig;/* MNF trigger loc. */ /* * These are based on the vmbus_chanmsg_choffer.chm_chanid. * Save it here for easy access. */ - volatile u_long *ch_evtflag;/* event flag loc. */ u_long ch_evtflag_mask;/* event flag */ + volatile u_long *ch_evtflag;/* event flag loc. */ /* -* TX bufring; at the beginning of ch_bufring. -*/ - hv_vmbus_ring_buffer_info ch_txbr; - /* -* RX bufring; immediately following ch_txbr. +* Rarely used fields. */ - hv_vmbus_ring_buffer_info ch_rxbr; - - struct taskqueue*ch_tq; - struct task ch_task; - vmbus_chan_callback_t ch_cb; - void*ch_cbarg; struct hyperv_mon_param *ch_monprm; struct hyperv_dma ch_monprm_dma; + uint32_tch_id; /* channel id */ + device_tch_dev; + struct vmbus_softc *ch_vmbus; +
svn commit: r303129 - in head/sys/dev/hyperv: include vmbus
Author: sephe Date: Thu Jul 21 05:46:26 2016 New Revision: 303129 URL: https://svnweb.freebsd.org/changeset/base/303129 Log: hyperv/vmbus: Shuffle function declaration and macro definition. MFC after:1 week Sponsored by: Microsoft OSTC Differential Revision:https://reviews.freebsd.org/D7259 Modified: head/sys/dev/hyperv/include/hyperv.h head/sys/dev/hyperv/include/vmbus.h head/sys/dev/hyperv/vmbus/vmbus_chanvar.h head/sys/dev/hyperv/vmbus/vmbus_var.h Modified: head/sys/dev/hyperv/include/hyperv.h == --- head/sys/dev/hyperv/include/hyperv.hThu Jul 21 05:38:05 2016 (r303128) +++ head/sys/dev/hyperv/include/hyperv.hThu Jul 21 05:46:26 2016 (r303129) @@ -58,23 +58,6 @@ #include #include -/* - * VMBUS version is 32 bit, upper 16 bit for major_number and lower - * 16 bit for minor_number. - * - * 0.13 -- Windows Server 2008 - * 1.1 -- Windows 7 - * 2.4 -- Windows 8 - * 3.0 -- Windows 8.1 - */ -#define VMBUS_VERSION_WS2008 ((0 << 16) | (13)) -#define VMBUS_VERSION_WIN7 ((1 << 16) | (1)) -#define VMBUS_VERSION_WIN8 ((2 << 16) | (4)) -#define VMBUS_VERSION_WIN8_1 ((3 << 16) | (0)) - -#define VMBUS_VERSION_MAJOR(ver) (((uint32_t)(ver)) >> 16) -#define VMBUS_VERSION_MINOR(ver) (((uint32_t)(ver)) & 0x) - struct hyperv_guid { uint8_t hv_guid[16]; } __packed; @@ -83,8 +66,6 @@ struct hyperv_guid { inthyperv_guid2str(const struct hyperv_guid *, char *, size_t); -struct hv_vmbus_channel; - /** * @brief Get physical address from virtual */ @@ -96,10 +77,4 @@ hv_get_phys_addr(void *virt) return (ret); } -static __inline struct hv_vmbus_channel * -vmbus_get_channel(device_t dev) -{ - return device_get_ivars(dev); -} - #endif /* __HYPERV_H__ */ Modified: head/sys/dev/hyperv/include/vmbus.h == --- head/sys/dev/hyperv/include/vmbus.h Thu Jul 21 05:38:05 2016 (r303128) +++ head/sys/dev/hyperv/include/vmbus.h Thu Jul 21 05:46:26 2016 (r303129) @@ -32,6 +32,23 @@ #include /* + * VMBUS version is 32 bit, upper 16 bit for major_number and lower + * 16 bit for minor_number. + * + * 0.13 -- Windows Server 2008 + * 1.1 -- Windows 7 + * 2.4 -- Windows 8 + * 3.0 -- Windows 8.1 + */ +#define VMBUS_VERSION_WS2008 ((0 << 16) | (13)) +#define VMBUS_VERSION_WIN7 ((1 << 16) | (1)) +#define VMBUS_VERSION_WIN8 ((2 << 16) | (4)) +#define VMBUS_VERSION_WIN8_1 ((3 << 16) | (0)) + +#define VMBUS_VERSION_MAJOR(ver) (((uint32_t)(ver)) >> 16) +#define VMBUS_VERSION_MINOR(ver) (((uint32_t)(ver)) & 0x) + +/* * GPA stuffs. */ struct vmbus_gpa_range { @@ -92,6 +109,12 @@ struct hyperv_guid; typedef void (*vmbus_chan_callback_t)(struct hv_vmbus_channel *, void *); +static __inline struct hv_vmbus_channel * +vmbus_get_channel(device_t dev) +{ + return device_get_ivars(dev); +} + intvmbus_chan_open(struct hv_vmbus_channel *chan, int txbr_size, int rxbr_size, const void *udata, int udlen, vmbus_chan_callback_t cb, void *cbarg); Modified: head/sys/dev/hyperv/vmbus/vmbus_chanvar.h == --- head/sys/dev/hyperv/vmbus/vmbus_chanvar.h Thu Jul 21 05:38:05 2016 (r303128) +++ head/sys/dev/hyperv/vmbus/vmbus_chanvar.h Thu Jul 21 05:46:26 2016 (r303129) @@ -181,4 +181,12 @@ typedef struct hv_vmbus_channel { #define VMBUS_CHAN_ST_OPENED_SHIFT 0 #define VMBUS_CHAN_ST_OPENED (1 << VMBUS_CHAN_ST_OPENED_SHIFT) +struct vmbus_softc; +struct vmbus_message; + +void vmbus_event_proc(struct vmbus_softc *, int); +void vmbus_event_proc_compat(struct vmbus_softc *, int); +void vmbus_chan_msgproc(struct vmbus_softc *, const struct vmbus_message *); +void vmbus_chan_destroy_all(struct vmbus_softc *); + #endif /* !_VMBUS_CHANVAR_H_ */ Modified: head/sys/dev/hyperv/vmbus/vmbus_var.h == --- head/sys/dev/hyperv/vmbus/vmbus_var.h Thu Jul 21 05:38:05 2016 (r303128) +++ head/sys/dev/hyperv/vmbus/vmbus_var.h Thu Jul 21 05:46:26 2016 (r303129) @@ -143,16 +143,11 @@ struct trapframe; struct vmbus_message; struct vmbus_msghc; -void vmbus_event_proc(struct vmbus_softc *, int); -void vmbus_event_proc_compat(struct vmbus_softc *, int); void vmbus_handle_intr(struct trapframe *); intvmbus_add_child(struct hv_vmbus_channel *); intvmbus_delete_child(struct hv_vmbus_channel *); - void vmbus_et_intr(struct trapframe *); - -void vmbus_chan_msgproc(struct vmbus_softc *, const struct vmbus_message *); -void vmbus_chan_destroy_all(struct vmbus_softc *); +uint32_t vmbus_gpadl_alloc(s
svn commit: r303130 - in head: . share/man/man7
Author: bdrewery Date: Thu Jul 21 06:06:11 2016 New Revision: 303130 URL: https://svnweb.freebsd.org/changeset/base/303130 Log: Add LOCAL_XTOOL_DIRS to add into cross-tools. Sponsored by: EMC / Isilon Storage Division Modified: head/Makefile.inc1 head/share/man/man7/build.7 Modified: head/Makefile.inc1 == --- head/Makefile.inc1 Thu Jul 21 05:46:26 2016(r303129) +++ head/Makefile.inc1 Thu Jul 21 06:06:11 2016(r303130) @@ -22,6 +22,8 @@ # to be created before files are installed # LOCAL_TOOL_DIRS="list of dirs" to add additional dirs to the build-tools # list +# LOCAL_XTOOL_DIRS="list of dirs" to add additional dirs to the +# cross-tools target # METALOG="path to metadata log" to write permission and ownership # when NO_ROOT is set. (default: ${DESTDIR}/METALOG) # TARGET="machine" to crossbuild world for a different machine type @@ -1815,6 +1817,7 @@ _usb_tools= sys/boot/usb/tools cross-tools: .MAKE .PHONY .for _tool in \ +${LOCAL_XTOOL_DIRS} \ ${_clang_libs} \ ${_clang} \ ${_binutils} \ Modified: head/share/man/man7/build.7 == --- head/share/man/man7/build.7 Thu Jul 21 05:46:26 2016(r303129) +++ head/share/man/man7/build.7 Thu Jul 21 06:06:11 2016(r303130) @@ -24,7 +24,7 @@ .\" .\" $FreeBSD$ .\" -.Dd April 28, 2016 +.Dd July 20, 2016 .Dt BUILD 7 .Os .Sh NAME @@ -462,6 +462,11 @@ If set, this variable supplies a list of the root of the source tree to build as part of the .Cm build-tools target. +.It Va LOCAL_XTOOL_DIRS +If set, this variable supplies a list of additional directories relative to +the root of the source tree to build as part of the +.Cm cross-tools +target. .It Va PORTS_MODULES A list of ports with kernel modules that should be built and installed as part of the ___ 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: r303131 - in head/sys/dev/hyperv: include netvsc storvsc utilities vmbus
Author: sephe Date: Thu Jul 21 06:09:47 2016 New Revision: 303131 URL: https://svnweb.freebsd.org/changeset/base/303131 Log: hyperv: hv_vmbus_channel -> vmbus_channel MFC after:1 week Sponsored by: Microsoft OSTC Differential Revision:https://reviews.freebsd.org/D7260 Modified: head/sys/dev/hyperv/include/vmbus.h head/sys/dev/hyperv/netvsc/hv_net_vsc.c head/sys/dev/hyperv/netvsc/hv_net_vsc.h head/sys/dev/hyperv/netvsc/hv_netvsc_drv_freebsd.c head/sys/dev/hyperv/netvsc/hv_rndis.h head/sys/dev/hyperv/netvsc/hv_rndis_filter.c head/sys/dev/hyperv/netvsc/hv_rndis_filter.h head/sys/dev/hyperv/storvsc/hv_storvsc_drv_freebsd.c head/sys/dev/hyperv/utilities/hv_heartbeat.c head/sys/dev/hyperv/utilities/hv_kvp.c head/sys/dev/hyperv/utilities/hv_shutdown.c head/sys/dev/hyperv/utilities/hv_timesync.c head/sys/dev/hyperv/utilities/hv_util.c head/sys/dev/hyperv/utilities/hv_util.h head/sys/dev/hyperv/vmbus/vmbus.c head/sys/dev/hyperv/vmbus/vmbus_chan.c head/sys/dev/hyperv/vmbus/vmbus_chanvar.h head/sys/dev/hyperv/vmbus/vmbus_var.h Modified: head/sys/dev/hyperv/include/vmbus.h == --- head/sys/dev/hyperv/include/vmbus.h Thu Jul 21 06:06:11 2016 (r303130) +++ head/sys/dev/hyperv/include/vmbus.h Thu Jul 21 06:09:47 2016 (r303131) @@ -104,56 +104,56 @@ struct vmbus_chanpkt_rxbuf { #define VMBUS_CHAN_SGLIST_MAX 32 #define VMBUS_CHAN_PRPLIST_MAX 32 -struct hv_vmbus_channel; +struct vmbus_channel; struct hyperv_guid; -typedef void (*vmbus_chan_callback_t)(struct hv_vmbus_channel *, void *); +typedef void (*vmbus_chan_callback_t)(struct vmbus_channel *, void *); -static __inline struct hv_vmbus_channel * +static __inline struct vmbus_channel * vmbus_get_channel(device_t dev) { return device_get_ivars(dev); } -intvmbus_chan_open(struct hv_vmbus_channel *chan, +intvmbus_chan_open(struct vmbus_channel *chan, int txbr_size, int rxbr_size, const void *udata, int udlen, vmbus_chan_callback_t cb, void *cbarg); -void vmbus_chan_close(struct hv_vmbus_channel *chan); +void vmbus_chan_close(struct vmbus_channel *chan); -intvmbus_chan_gpadl_connect(struct hv_vmbus_channel *chan, +intvmbus_chan_gpadl_connect(struct vmbus_channel *chan, bus_addr_t paddr, int size, uint32_t *gpadl); -intvmbus_chan_gpadl_disconnect(struct hv_vmbus_channel *chan, +intvmbus_chan_gpadl_disconnect(struct vmbus_channel *chan, uint32_t gpadl); -void vmbus_chan_cpu_set(struct hv_vmbus_channel *chan, int cpu); -void vmbus_chan_cpu_rr(struct hv_vmbus_channel *chan); -struct hv_vmbus_channel * - vmbus_chan_cpu2chan(struct hv_vmbus_channel *chan, int cpu); -void vmbus_chan_set_readbatch(struct hv_vmbus_channel *chan, bool on); - -struct hv_vmbus_channel ** - vmbus_subchan_get(struct hv_vmbus_channel *pri_chan, int subchan_cnt); -void vmbus_subchan_rel(struct hv_vmbus_channel **subchan, int subchan_cnt); -void vmbus_subchan_drain(struct hv_vmbus_channel *pri_chan); +void vmbus_chan_cpu_set(struct vmbus_channel *chan, int cpu); +void vmbus_chan_cpu_rr(struct vmbus_channel *chan); +struct vmbus_channel * + vmbus_chan_cpu2chan(struct vmbus_channel *chan, int cpu); +void vmbus_chan_set_readbatch(struct vmbus_channel *chan, bool on); + +struct vmbus_channel ** + vmbus_subchan_get(struct vmbus_channel *pri_chan, int subchan_cnt); +void vmbus_subchan_rel(struct vmbus_channel **subchan, int subchan_cnt); +void vmbus_subchan_drain(struct vmbus_channel *pri_chan); -intvmbus_chan_recv(struct hv_vmbus_channel *chan, void *data, int *dlen, +intvmbus_chan_recv(struct vmbus_channel *chan, void *data, int *dlen, uint64_t *xactid); -intvmbus_chan_recv_pkt(struct hv_vmbus_channel *chan, +intvmbus_chan_recv_pkt(struct vmbus_channel *chan, struct vmbus_chanpkt_hdr *pkt, int *pktlen); -intvmbus_chan_send(struct hv_vmbus_channel *chan, uint16_t type, +intvmbus_chan_send(struct vmbus_channel *chan, uint16_t type, uint16_t flags, void *data, int dlen, uint64_t xactid); -intvmbus_chan_send_sglist(struct hv_vmbus_channel *chan, +intvmbus_chan_send_sglist(struct vmbus_channel *chan, struct vmbus_gpa sg[], int sglen, void *data, int dlen, uint64_t xactid); -intvmbus_chan_send_prplist(struct hv_vmbus_channel *chan, +intvmbus_chan_send_prplist(struct vmbus_channel *chan, struct vmbus_gpa_range *prp, int prp_cnt, void *data, int dlen, uint64_t xactid); -uint32_t vmbus_chan_id(const struct hv_vmbus_channel *chan); -uint32_t vmbus_chan_subidx(const struct hv_vmbus_channel *chan); -bool vmbus_chan_is_primary(const struct hv_vmbus_channel *chan); +uint32_t vmbus_chan_id(const struct vmbus_channel *chan); +uint32_t vmbus_chan_subidx(const struct vmbus_channel *chan