Re: svn commit: r304510 - head/sys/dev/ixl

2016-08-31 Thread Eric Joyner 
It's weird that the gcc message is mentioning "ixlv_media_change", but I'll
do something about it shortly.

On Wed, Aug 31, 2016 at 2:16 PM Dimitry Andric  wrote:

> On 31 Aug 2016, at 18:53, Ed Maste  wrote:
> >
> > On 19 August 2016 at 20:08, Eric Joyner  wrote:
> >> Author: erj
> >> Date: Sat Aug 20 00:08:10 2016
> >> New Revision: 304510
> >> URL: https://svnweb.freebsd.org/changeset/base/304510
> >>
> >> Log:
> >>  ixlv(4): Fix ixlv(4) not loading when loaded as a kernel module and
> netmap is enabled.
> >
> > I did not confirm it's this commit, but kernel builds currently have
> > this warning:
> >
> > if_ixlv.o: warning: common of `ixl_crcstrip' overridden by definition
> > if_ixl.o: warning: defined here
> > if_ixlv.o: warning: multiple common of `ixl_rx_miss'
> > if_ixl.o: warning: previous common is here
> > if_ixlv.o: warning: multiple common of `ixl_rx_miss_bufs'
> > if_ixl.o: warning: previous common is here
>
> Interestingly, with the external gcc build, it is even an error:
>
> --- kernel.full ---
> linking kernel.full
> if_ixlv.o:(.bss+0x0): multiple definition of `ixl_crcstrip'
> if_ixl.o:(.data+0xb8): first defined here
> if_ixlv.o: In function `ixlv_media_change':
> /builds/FreeBSD_HEAD_amd64_gcc/sys/dev/ixl/if_ixlv.c:2252: multiple
> definition of `ixl_rx_miss_bufs'
> if_ixl.o:/builds/FreeBSD_HEAD_amd64_gcc/sys/dev/ixl/if_ixl.c:680: first
> defined here
> if_ixlv.o: In function `ixlv_media_change':
> /builds/FreeBSD_HEAD_amd64_gcc/sys/dev/ixl/if_ixlv.c:2262: multiple
> definition of `ixl_rx_miss'
> if_ixl.o:/builds/FreeBSD_HEAD_amd64_gcc/sys/dev/ixl/if_ixl.c:681: first
> defined here
> *** [kernel.full] Error code 1
>
> See e.g.
> https://jenkins.freebsd.org/job/FreeBSD_HEAD_amd64_gcc/1505/console
>
> -Dimitry
>
>
___
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: r305171 - head/sys/kern

2016-08-31 Thread Ed Maste 
Author: emaste
Date: Thu Sep  1 02:51:50 2016
New Revision: 305171
URL: https://svnweb.freebsd.org/changeset/base/305171

Log:
  allow kern.proc.nfds sysctl in capability mode
  
  Reviewed by:  allanjude
  MFC after:1 week
  Sponsored by: The FreeBSD Foundation
  Differential Revision:https://reviews.freebsd.org/D7733

Modified:
  head/sys/kern/kern_descrip.c

Modified: head/sys/kern/kern_descrip.c
==
--- head/sys/kern/kern_descrip.cThu Sep  1 02:42:11 2016
(r305170)
+++ head/sys/kern/kern_descrip.cThu Sep  1 02:51:50 2016
(r305171)
@@ -3178,7 +3178,7 @@ sysctl_kern_proc_nfds(SYSCTL_HANDLER_ARG
 }
 
 static SYSCTL_NODE(_kern_proc, KERN_PROC_NFDS, nfds,
-CTLFLAG_RD|CTLFLAG_MPSAFE, sysctl_kern_proc_nfds,
+CTLFLAG_RD|CTLFLAG_CAPRD|CTLFLAG_MPSAFE, sysctl_kern_proc_nfds,
 "Number of open file descriptors");
 
 /*
___
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: r305170 - head/lib/atf/libatf-c

2016-08-31 Thread Ngie Cooper 
Author: ngie
Date: Thu Sep  1 02:42:11 2016
New Revision: 305170
URL: https://svnweb.freebsd.org/changeset/base/305170

Log:
  Don't bake all of CC/CPP/CXX into CFLAGS
  
  Capture executable names for CC, CPP, CXX (assumed to be the
  first non-CCACHE_BIN word).
  
  This change strips out all of the cross-compiler arguments, (-target,
  -B, etc), added to ${CC}, etc via ${CROSSENV} in Makefile.inc1, so it
  doesn't infect the build and subsequently the test.
  
  Add comments noting why this logic is being added, and why the logic in
  r305041 was necessary/what it was trying to achieve.
  
  This is required after recent changes made to the toolchain to always
  specify --sysroot, -target, -B, etc with clang in buildworld (presumably
  r304681).
  
  Reviewed by:  rodrigc (earlier version)
  Reported by:  Jenkins (FreeBSD_HEAD job from 559+)
  MFC after:12 days
  X-MFC with:   r304681, r305041
  Sponsored by: EMC / Isilon Storage Division
  Differential Revision:https://reviews.freebsd.org/D7732

Modified:
  head/lib/atf/libatf-c/Makefile

Modified: head/lib/atf/libatf-c/Makefile
==
--- head/lib/atf/libatf-c/Makefile  Thu Sep  1 02:05:46 2016
(r305169)
+++ head/lib/atf/libatf-c/Makefile  Thu Sep  1 02:42:11 2016
(r305170)
@@ -28,6 +28,18 @@
 .include 
 .include 
 
+# Store the toolchain executable in ATF_BUILD_{CC,CPP,CXX} to ensure other
+# values -- like -target, -B ..., etc -- don't get leaked into the tests.
+#
+# Be sure to omit ${CCACHE_BIN} (if specified) from the variable as it gets
+# automatically appended to the variables in bsd.compiler.mk when
+# ${MK_CCACHE_BUILD} != no.
+ATF_BUILD_CC:= ${CC:N${CCACHE_BIN}:[1]}
+ATF_BUILD_CPP:=${CPP:N${CCACHE_BIN}:[1]}
+ATF_BUILD_CXX:=${CXX:N${CCACHE_BIN}:[1]}
+
+# Only capture defines, includes, linker flags, optimization levels, warnings
+# and preprocessor flags when building ATF_BUILD_{C,CPP,CXX}FLAGS.
 ATF_BUILD_CFLAGS:= ${CFLAGS:M-[DILOWf]*}
 ATF_BUILD_CPPFLAGS:=   ${CPPFLAGS:M-[DILOWf]*}
 ATF_BUILD_CXXFLAGS:=   ${CXXFLAGS:M-[DILOWf]*}
@@ -41,11 +53,11 @@ ATF=${SRCTOP}/contrib/atf
 .PATH: ${ATF}/atf-c
 .PATH: ${ATF}/atf-c/detail
 
-CFLAGS+=   -DATF_BUILD_CC='"${CC}"'
+CFLAGS+=   -DATF_BUILD_CC='"${ATF_BUILD_CC}"'
 CFLAGS+=   -DATF_BUILD_CFLAGS='"${ATF_BUILD_CFLAGS}"'
-CFLAGS+=   -DATF_BUILD_CPP='"${CPP}"'
+CFLAGS+=   -DATF_BUILD_CPP='"${ATF_BUILD_CPP}"'
 CFLAGS+=   -DATF_BUILD_CPPFLAGS='"${ATF_BUILD_CPPFLAGS}"'
-CFLAGS+=   -DATF_BUILD_CXX='"${CXX}"'
+CFLAGS+=   -DATF_BUILD_CXX='"${ATF_BUILD_CXX}"'
 CFLAGS+=   -DATF_BUILD_CXXFLAGS='"${ATF_BUILD_CXXFLAGS}"'
 CFLAGS+=   -I${ATF}
 CFLAGS+=   -I${.CURDIR}
___
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: r301752 - head/contrib/netbsd-tests/lib/libc/string

2016-08-31 Thread Ngie Cooper (yaneurabeya) 

> On Aug 31, 2016, at 7:11 PM, Craig Rodrigues  wrote:
> 
> Ngie,
> 
> Can you please MFC this to stable/10 branch?
> I believe this has been causing a test failure in this branch
> since June.

Shoot — I’ll verify the issue and commit the change.
Thanks,
-Ngie

___
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: r301752 - head/contrib/netbsd-tests/lib/libc/string

2016-08-31 Thread Craig Rodrigues 
Ngie,

Can you please MFC this to stable/10 branch?
I believe this has been causing a test failure in this branch
since June.

Thanks.

--
Craig


On Thu, Jun 9, 2016 at 11:35 AM, Garrett Cooper  wrote:

> Author: ngie
> Date: Thu Jun  9 18:35:37 2016
> New Revision: 301752
> URL: https://svnweb.freebsd.org/changeset/base/301752
>
> Log:
>   Update `goodResult` after recent changes made to the PRNG in libc
>
>   The PRNG was changed in r300953/r300956, and subsequently, the numbers
>   generated have changed. This is expected ABI breakage per ache
>
>   X-MFC with: r300953, r300956
>   Tested with: amd64, i386
>   Sponsored by: EMC / Isilon Storage Division
>
> Modified:
>   head/contrib/netbsd-tests/lib/libc/string/t_memcpy.c
>
> Modified: head/contrib/netbsd-tests/lib/libc/string/t_memcpy.c
> 
> ==
> --- head/contrib/netbsd-tests/lib/libc/string/t_memcpy.cThu Jun
> 9 18:27:41 2016(r301751)
> +++ head/contrib/netbsd-tests/lib/libc/string/t_memcpy.cThu Jun
> 9 18:35:37 2016(r301752)
> @@ -54,7 +54,7 @@ char result[100];
>  #ifdef __NetBSD__
>  const char goodResult[] = "7b405d24bc03195474c70ddae9e1f8fb";
>  #else
> -const char goodResult[] = "217b4fbe456916bf62a2f85df752e4ab";
> +const char goodResult[] = "5ab4443f0e3e058d94087d9f2a11ef5e";
>  #endif
>
>  static void
>
>
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"

svn commit: r305169 - head/sys/kern

2016-08-31 Thread Patrick Kelsey 
Author: pkelsey
Date: Thu Sep  1 02:05:46 2016
New Revision: 305169
URL: https://svnweb.freebsd.org/changeset/base/305169

Log:
  _taskqueue_start_threads() now fails if it doesn't actually start any threads.
  
  Reviewed by:  jhb
  MFC after:1 week
  Differential Revision:https://reviews.freebsd.org/D7701

Modified:
  head/sys/kern/subr_taskqueue.c

Modified: head/sys/kern/subr_taskqueue.c
==
--- head/sys/kern/subr_taskqueue.c  Thu Sep  1 01:08:18 2016
(r305168)
+++ head/sys/kern/subr_taskqueue.c  Thu Sep  1 02:05:46 2016
(r305169)
@@ -625,6 +625,11 @@ _taskqueue_start_threads(struct taskqueu
} else
tq->tq_tcount++;
}
+   if (tq->tq_tcount == 0) {
+   free(tq->tq_threads, M_TASKQUEUE);
+   tq->tq_threads = NULL;
+   return (ENOMEM);
+   }
for (i = 0; i < count; i++) {
if (tq->tq_threads[i] == 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: r305168 - in head/sys/dev: ixl netmap

2016-08-31 Thread Eric Joyner 
Author: erj
Date: Thu Sep  1 01:08:18 2016
New Revision: 305168
URL: https://svnweb.freebsd.org/changeset/base/305168

Log:
  Fix linker warnings (errors on gcc) that resulted from r304510.
  
  The variables that are extern in the netmap header file should be
  defined in ixl_txrx.c (the file that is included in both ixl(4)/ixlv(4),
  not in the main driver source files.
  
  Reported by:  ed@, dim@, ngie@

Modified:
  head/sys/dev/ixl/if_ixlv.c
  head/sys/dev/ixl/ixl_txrx.c
  head/sys/dev/netmap/if_ixl_netmap.h

Modified: head/sys/dev/ixl/if_ixlv.c
==
--- head/sys/dev/ixl/if_ixlv.c  Thu Sep  1 00:51:59 2016(r305167)
+++ head/sys/dev/ixl/if_ixlv.c  Thu Sep  1 01:08:18 2016(r305168)
@@ -217,11 +217,6 @@ TUNABLE_INT("hw.ixlv.tx_itr", _tx_i
 SYSCTL_INT(_hw_ixlv, OID_AUTO, tx_itr, CTLFLAG_RDTUN,
 _tx_itr, 0, "TX Interrupt Rate");
 
-/* Fix when building as a standalone module when netmap is enabled */
-#if defined(DEV_NETMAP) && !defined(NETMAP_IXL_MAIN)
-int ixl_rx_miss, ixl_rx_miss_bufs, ixl_crcstrip;
-#endif
-
 /*
  *  Device identification routine
  *

Modified: head/sys/dev/ixl/ixl_txrx.c
==
--- head/sys/dev/ixl/ixl_txrx.c Thu Sep  1 00:51:59 2016(r305167)
+++ head/sys/dev/ixl/ixl_txrx.c Thu Sep  1 01:08:18 2016(r305168)
@@ -69,6 +69,7 @@ static inline u32 ixl_get_tx_head(struct
 
 #ifdef DEV_NETMAP
 #include 
+int ixl_rx_miss, ixl_rx_miss_bufs, ixl_crcstrip = 1;
 #endif /* DEV_NETMAP */
 
 /*

Modified: head/sys/dev/netmap/if_ixl_netmap.h
==
--- head/sys/dev/netmap/if_ixl_netmap.h Thu Sep  1 00:51:59 2016
(r305167)
+++ head/sys/dev/netmap/if_ixl_netmap.h Thu Sep  1 01:08:18 2016
(r305168)
@@ -71,7 +71,6 @@ SYSCTL_DECL(_dev_netmap);
 /*
  * The xl driver by default strips CRCs and we do not override it.
  */
-int ixl_rx_miss, ixl_rx_miss_bufs, ixl_crcstrip = 1;
 #if 0
 SYSCTL_INT(_dev_netmap, OID_AUTO, ixl_crcstrip,
 CTLFLAG_RW, _crcstrip, 1, "strip CRC on rx frames");
___
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: r305167 - head/sys/dev/cxgbe/tom

2016-08-31 Thread Navdeep Parhar 
Author: np
Date: Thu Sep  1 00:51:59 2016
New Revision: 305167
URL: https://svnweb.freebsd.org/changeset/base/305167

Log:
  cxgbe/t4_tom: Two new routines to allocate and write page pods for a
  buffer in the kernel's address space.

Modified:
  head/sys/dev/cxgbe/tom/t4_ddp.c
  head/sys/dev/cxgbe/tom/t4_tom.h

Modified: head/sys/dev/cxgbe/tom/t4_ddp.c
==
--- head/sys/dev/cxgbe/tom/t4_ddp.c Wed Aug 31 23:23:46 2016
(r305166)
+++ head/sys/dev/cxgbe/tom/t4_ddp.c Thu Sep  1 00:51:59 2016
(r305167)
@@ -936,6 +936,64 @@ have_pgsz:
return (1);
 }
 
+int
+t4_alloc_page_pods_for_buf(struct ppod_region *pr, vm_offset_t buf, int len,
+struct ppod_reservation *prsv)
+{
+   int hcf, seglen, idx, npages, nppods;
+   uintptr_t start_pva, end_pva, pva, p1;
+
+   MPASS(buf > 0);
+   MPASS(len > 0);
+
+   /*
+* The DDP page size is unrelated to the VM page size.  We combine
+* contiguous physical pages into larger segments to get the best DDP
+* page size possible.  This is the largest of the four sizes in
+* A_ULP_RX_ISCSI_PSZ that evenly divides the HCF of the segment sizes
+* in the page list.
+*/
+   hcf = 0;
+   start_pva = trunc_page(buf);
+   end_pva = trunc_page(buf + len - 1);
+   pva = start_pva;
+   while (pva <= end_pva) {
+   seglen = PAGE_SIZE;
+   p1 = pmap_kextract(pva);
+   pva += PAGE_SIZE;
+   while (pva <= end_pva && p1 + seglen == pmap_kextract(pva)) {
+   seglen += PAGE_SIZE;
+   pva += PAGE_SIZE;
+   }
+
+   hcf = calculate_hcf(hcf, seglen);
+   if (hcf < (1 << pr->pr_page_shift[1])) {
+   idx = 0;
+   goto have_pgsz; /* give up, short circuit */
+   }
+   }
+
+#define PR_PAGE_MASK(x) ((1 << pr->pr_page_shift[(x)]) - 1)
+   MPASS((hcf & PR_PAGE_MASK(0)) == 0); /* PAGE_SIZE is >= 4K everywhere */
+   for (idx = nitems(pr->pr_page_shift) - 1; idx > 0; idx--) {
+   if ((hcf & PR_PAGE_MASK(idx)) == 0)
+   break;
+   }
+#undef PR_PAGE_MASK
+
+have_pgsz:
+   MPASS(idx <= M_PPOD_PGSZ);
+
+   npages = 1;
+   npages += (end_pva - start_pva) >> pr->pr_page_shift[idx];
+   nppods = howmany(npages, PPOD_PAGES);
+   if (alloc_page_pods(pr, nppods, idx, prsv) != 0)
+   return (ENOMEM);
+   MPASS(prsv->prsv_nppods > 0);
+
+   return (0);
+}
+
 void
 t4_free_page_pods(struct ppod_reservation *prsv)
 {
@@ -1036,6 +1094,94 @@ t4_write_page_pods_for_ps(struct adapter
return (0);
 }
 
+int
+t4_write_page_pods_for_buf(struct adapter *sc, struct sge_wrq *wrq, int tid,
+struct ppod_reservation *prsv, vm_offset_t buf, int buflen)
+{
+   struct wrqe *wr;
+   struct ulp_mem_io *ulpmc;
+   struct ulptx_idata *ulpsc;
+   struct pagepod *ppod;
+   int i, j, k, n, chunk, len, ddp_pgsz;
+   u_int ppod_addr, offset;
+   uint32_t cmd;
+   struct ppod_region *pr = prsv->prsv_pr;
+   uintptr_t end_pva, pva, pa;
+
+   cmd = htobe32(V_ULPTX_CMD(ULP_TX_MEM_WRITE));
+   if (is_t4(sc))
+   cmd |= htobe32(F_ULP_MEMIO_ORDER);
+   else
+   cmd |= htobe32(F_T5_ULP_MEMIO_IMM);
+   ddp_pgsz = 1 << pr->pr_page_shift[G_PPOD_PGSZ(prsv->prsv_tag)];
+   offset = buf & PAGE_MASK;
+   ppod_addr = pr->pr_start + (prsv->prsv_tag & pr->pr_tag_mask);
+   pva = trunc_page(buf);
+   end_pva = trunc_page(buf + buflen - 1);
+   for (i = 0; i < prsv->prsv_nppods; ppod_addr += chunk) {
+
+   /* How many page pods are we writing in this cycle */
+   n = min(prsv->prsv_nppods - i, NUM_ULP_TX_SC_IMM_PPODS);
+   MPASS(n > 0);
+   chunk = PPOD_SZ(n);
+   len = roundup2(sizeof(*ulpmc) + sizeof(*ulpsc) + chunk, 16);
+
+   wr = alloc_wrqe(len, wrq);
+   if (wr == NULL)
+   return (ENOMEM);/* ok to just bail out */
+   ulpmc = wrtod(wr);
+
+   INIT_ULPTX_WR(ulpmc, len, 0, 0);
+   ulpmc->cmd = cmd;
+   ulpmc->dlen = htobe32(V_ULP_MEMIO_DATA_LEN(chunk / 32));
+   ulpmc->len16 = htobe32(howmany(len - sizeof(ulpmc->wr), 16));
+   ulpmc->lock_addr = htobe32(V_ULP_MEMIO_ADDR(ppod_addr >> 5));
+
+   ulpsc = (struct ulptx_idata *)(ulpmc + 1);
+   ulpsc->cmd_more = htobe32(V_ULPTX_CMD(ULP_TX_SC_IMM));
+   ulpsc->len = htobe32(chunk);
+
+   ppod = (struct pagepod *)(ulpsc + 1);
+   for (j = 0; j < n; i++, j++, ppod++) {
+   ppod->vld_tid_pgsz_tag_color = htobe64(F_PPOD_VALID |
+   V_PPOD_TID(tid) |
+

svn commit: r305166 - in head/sys/dev/cxgbe: cxgbei tom

2016-08-31 Thread Navdeep Parhar 
Author: np
Date: Wed Aug 31 23:23:46 2016
New Revision: 305166
URL: https://svnweb.freebsd.org/changeset/base/305166

Log:
  cxgbe/t4_tom: Add general purpose routines to deal with page pod regions
  and allocations within them.  Switch to these routines to manage the TOE
  DDP region.
  
  Sponsored by: Chelsio Communications

Modified:
  head/sys/dev/cxgbe/cxgbei/cxgbei.c
  head/sys/dev/cxgbe/tom/t4_ddp.c
  head/sys/dev/cxgbe/tom/t4_tom.c
  head/sys/dev/cxgbe/tom/t4_tom.h

Modified: head/sys/dev/cxgbe/cxgbei/cxgbei.c
==
--- head/sys/dev/cxgbe/cxgbei/cxgbei.c  Wed Aug 31 22:18:50 2016
(r305165)
+++ head/sys/dev/cxgbe/cxgbei/cxgbei.c  Wed Aug 31 23:23:46 2016
(r305166)
@@ -157,7 +157,6 @@ ulp_mem_io_set_hdr(struct adapter *sc, i
idata->len = htonl(dlen);
 }
 
-#define PPOD_SIZE  sizeof(struct pagepod)
 #define ULPMEM_IDATA_MAX_NPPODS 1  /* 256/PPOD_SIZE */
 #define PCIE_MEMWIN_MAX_NPPODS 16  /* 1024/PPOD_SIZE */
 

Modified: head/sys/dev/cxgbe/tom/t4_ddp.c
==
--- head/sys/dev/cxgbe/tom/t4_ddp.c Wed Aug 31 22:18:50 2016
(r305165)
+++ head/sys/dev/cxgbe/tom/t4_ddp.c Wed Aug 31 23:23:46 2016
(r305166)
@@ -85,44 +85,11 @@ static void ddp_complete_all(struct toep
 static void t4_aio_cancel_active(struct kaiocb *job);
 static void t4_aio_cancel_queued(struct kaiocb *job);
 
-#define PPOD_SZ(n) ((n) * sizeof(struct pagepod))
-#define PPOD_SIZE  (PPOD_SZ(1))
-
 static TAILQ_HEAD(, pageset) ddp_orphan_pagesets;
 static struct mtx ddp_orphan_pagesets_lock;
 static struct task ddp_orphan_task;
 
 #define MAX_DDP_BUFFER_SIZE(M_TCB_RX_DDP_BUF0_LEN)
-static int
-alloc_ppods(struct tom_data *td, int n, u_int *ppod_addr)
-{
-   vmem_addr_t v;
-   int rc;
-
-   MPASS(n > 0);
-
-   rc = vmem_alloc(td->ppod_arena, PPOD_SZ(n), M_NOWAIT | M_FIRSTFIT, );
-   *ppod_addr = (u_int)v;
-
-   return (rc);
-}
-
-static void
-free_ppods(struct tom_data *td, u_int ppod_addr, int n)
-{
-
-   MPASS(n > 0);
-
-   vmem_free(td->ppod_arena, (vmem_addr_t)ppod_addr, PPOD_SZ(n));
-}
-
-static inline int
-pages_to_nppods(int npages, int ddp_pgsz)
-{
-   int nsegs = npages * PAGE_SIZE / ddp_pgsz;
-
-   return (howmany(nsegs, PPOD_PAGES));
-}
 
 /*
  * A page set holds information about a buffer used for DDP.  The page
@@ -147,8 +114,8 @@ free_pageset(struct tom_data *td, struct
vm_page_t p;
int i;
 
-   if (ps->nppods > 0)
-   free_ppods(td, ps->ppod_addr, ps->nppods);
+   if (ps->prsv.prsv_nppods > 0)
+   t4_free_page_pods(>prsv);
 
if (ps->flags & PS_WIRED) {
for (i = 0; i < ps->npages; i++) {
@@ -483,7 +450,7 @@ mk_update_tcb_for_ddp(struct adapter *sc
ulpmc = mk_set_tcb_field_ulp(ulpmc, toep,
W_TCB_RX_DDP_BUF0_TAG + db_idx,
V_TCB_RX_DDP_BUF0_TAG(M_TCB_RX_DDP_BUF0_TAG),
-   V_TCB_RX_DDP_BUF0_TAG(ps->tag));
+   V_TCB_RX_DDP_BUF0_TAG(ps->prsv.prsv_tag));
 
/* Update the current offset in the DDP buffer and its total length */
if (db_idx == 0)
@@ -879,13 +846,52 @@ calculate_hcf(int n1, int n2)
return (b);
 }
 
+static inline int
+pages_to_nppods(int npages, int ddp_page_shift)
+{
+
+   MPASS(ddp_page_shift >= PAGE_SHIFT);
+
+   return (howmany(npages >> (ddp_page_shift - PAGE_SHIFT), PPOD_PAGES));
+}
+
 static int
-alloc_page_pods(struct tom_data *td, struct pageset *ps)
+alloc_page_pods(struct ppod_region *pr, u_int nppods, u_int pgsz_idx,
+struct ppod_reservation *prsv)
 {
-   int i, hcf, seglen, idx, ppod, nppods;
-   u_int ppod_addr;
+   vmem_addr_t addr;   /* relative to start of region */
+
+   if (vmem_alloc(pr->pr_arena, PPOD_SZ(nppods), M_NOWAIT | M_FIRSTFIT,
+   ) != 0)
+   return (ENOMEM);
+
+   CTR5(KTR_CXGBE, "%-17s arena %p, addr 0x%08x, nppods %d, pgsz %d",
+   __func__, pr->pr_arena, (uint32_t)addr & pr->pr_tag_mask,
+   nppods, 1 << pr->pr_page_shift[pgsz_idx]);
+
+   /*
+* The hardware tagmask includes an extra invalid bit but the arena was
+* seeded with valid values only.  An allocation out of this arena will
+* fit inside the tagmask but won't have the invalid bit set.
+*/
+   MPASS((addr & pr->pr_tag_mask) == addr);
+   MPASS((addr & pr->pr_invalid_bit) == 0);
+
+   prsv->prsv_pr = pr;
+   prsv->prsv_tag = V_PPOD_PGSZ(pgsz_idx) | addr;
+   prsv->prsv_nppods = nppods;
+
+   return (0);
+}
+
+int
+t4_alloc_page_pods_for_ps(struct ppod_region *pr, struct pageset *ps)
+{
+   int i, hcf, seglen, idx, nppods;
+   struct ppod_reservation *prsv = >prsv;
 
-   KASSERT(ps->nppods == 0, ("%s: page pods already allocated", __func__));
+   KASSERT(prsv->prsv_nppods == 0,
+

Re: svn commit: r304510 - head/sys/dev/ixl

2016-08-31 Thread Ngie Cooper 
On Wed, Aug 31, 2016 at 2:15 PM, Dimitry Andric  wrote:

...

> Interestingly, with the external gcc build, it is even an error:
>
> --- kernel.full ---
> linking kernel.full
> if_ixlv.o:(.bss+0x0): multiple definition of `ixl_crcstrip'
> if_ixl.o:(.data+0xb8): first defined here
> if_ixlv.o: In function `ixlv_media_change':
> /builds/FreeBSD_HEAD_amd64_gcc/sys/dev/ixl/if_ixlv.c:2252: multiple 
> definition of `ixl_rx_miss_bufs'
> if_ixl.o:/builds/FreeBSD_HEAD_amd64_gcc/sys/dev/ixl/if_ixl.c:680: first 
> defined here
> if_ixlv.o: In function `ixlv_media_change':
> /builds/FreeBSD_HEAD_amd64_gcc/sys/dev/ixl/if_ixlv.c:2262: multiple 
> definition of `ixl_rx_miss'
> if_ixl.o:/builds/FreeBSD_HEAD_amd64_gcc/sys/dev/ixl/if_ixl.c:681: first 
> defined here
> *** [kernel.full] Error code 1
>
> See e.g. https://jenkins.freebsd.org/job/FreeBSD_HEAD_amd64_gcc/1505/console

+1:

linking kernel.full
if_ixlv.o: warning: common of `ixl_crcstrip' overridden by definition
if_ixl.o: warning: defined here
if_ixlv.o: warning: multiple common of `ixl_rx_miss'
if_ixl.o: warning: previous common is here
if_ixlv.o: warning: multiple common of `ixl_rx_miss_bufs'
if_ixl.o: warning: previous common is here
___
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: r305165 - in stable: 10/sbin/resolvconf 11/sbin/resolvconf

2016-08-31 Thread Eric Badger 
Author: badger
Date: Wed Aug 31 22:18:50 2016
New Revision: 305165
URL: https://svnweb.freebsd.org/changeset/base/305165

Log:
  MFC r304652:
  
  Fix missing substitution of @SBINDIR@ in resolvconf scripts
  
  Certain features, such as resolv_conf_passthrough=NULL, do not work
  correctly due to this missing substitution.
  
  Also remove the @PREFIX@ substitution, which is no longer needed.
  
  Approved by:  vangyzen (mentor)
  Sponsored by: Dell Inc.

Modified:
  stable/10/sbin/resolvconf/Makefile
Directory Properties:
  stable/10/   (props changed)

Changes in other areas also in this revision:
Modified:
  stable/11/sbin/resolvconf/Makefile
Directory Properties:
  stable/11/   (props changed)

Modified: stable/10/sbin/resolvconf/Makefile
==
--- stable/10/sbin/resolvconf/Makefile  Wed Aug 31 21:48:22 2016
(r305164)
+++ stable/10/sbin/resolvconf/Makefile  Wed Aug 31 22:18:50 2016
(r305165)
@@ -15,6 +15,7 @@ CLEANFILES=   ${SCRIPTS} ${FILES} ${MAN}
 SYSCONFDIR=/etc
 RCDIR= ${SYSCONFDIR}/rc.d
 VARDIR=/var/run/resolvconf
+SBINDIR=   /sbin
 
 # We don't assume to restart the services in /sbin.  So, though
 # our service(8) is in /usr/sbin, we can use it, here.
@@ -24,12 +25,12 @@ RESTARTCMD= /usr/sbin/service ${CMD1} \&
 
 .for f in ${SCRIPTS} ${FILES} ${MAN}
 ${f}:  ${f}.in
-   sed -e 's:@PREFIX@::g' \
-   -e 's:@SYSCONFDIR@:${SYSCONFDIR}:g' \
+   sed -e 's:@SYSCONFDIR@:${SYSCONFDIR}:g' \
-e 's:@LIBEXECDIR@:${FILESDIR}:g' \
-e 's:@VARDIR@:${VARDIR}:g' \
-e 's:@RESTARTCMD \(.*\)@:${RESTARTCMD}:g' \
-e 's:@RCDIR@:${RCDIR}:g' \
+   -e 's:@SBINDIR@:${SBINDIR}:g' \
-e 's: vpn : ng[0-9]*&:g' \
${DIST}/$@.in > $@
 .endfor
___
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: r305165 - in stable: 10/sbin/resolvconf 11/sbin/resolvconf

2016-08-31 Thread Eric Badger 
Author: badger
Date: Wed Aug 31 22:18:50 2016
New Revision: 305165
URL: https://svnweb.freebsd.org/changeset/base/305165

Log:
  MFC r304652:
  
  Fix missing substitution of @SBINDIR@ in resolvconf scripts
  
  Certain features, such as resolv_conf_passthrough=NULL, do not work
  correctly due to this missing substitution.
  
  Also remove the @PREFIX@ substitution, which is no longer needed.
  
  Approved by:  vangyzen (mentor)
  Sponsored by: Dell Inc.

Modified:
  stable/11/sbin/resolvconf/Makefile
Directory Properties:
  stable/11/   (props changed)

Changes in other areas also in this revision:
Modified:
  stable/10/sbin/resolvconf/Makefile
Directory Properties:
  stable/10/   (props changed)

Modified: stable/11/sbin/resolvconf/Makefile
==
--- stable/11/sbin/resolvconf/Makefile  Wed Aug 31 21:48:22 2016
(r305164)
+++ stable/11/sbin/resolvconf/Makefile  Wed Aug 31 22:18:50 2016
(r305165)
@@ -16,6 +16,7 @@ CLEANFILES=   ${SCRIPTS} ${FILES} ${MAN}
 SYSCONFDIR=/etc
 RCDIR= ${SYSCONFDIR}/rc.d
 VARDIR=/var/run/resolvconf
+SBINDIR=   /sbin
 
 # We don't assume to restart the services in /sbin.  So, though
 # our service(8) is in /usr/sbin, we can use it, here.
@@ -28,13 +29,13 @@ RESTARTCMD= "/usr/sbin/service ${CMD1} \
 
 .for f in ${SCRIPTS} ${FILES} ${MAN}
 ${f}:  ${f}.in
-   sed -e 's:@PREFIX@::g' \
-   -e 's:@SYSCONFDIR@:${SYSCONFDIR}:g' \
+   sed -e 's:@SYSCONFDIR@:${SYSCONFDIR}:g' \
-e 's:@LIBEXECDIR@:${FILESDIR}:g' \
-e 's:@VARDIR@:${VARDIR}:g' \
-e 's:@RESTARTCMD \(.*\)@:${RESTARTCMD_WITH_ARG}:g' \
-e 's:@RESTARTCMD@:${RESTARTCMD}:g' \
-e 's:@RCDIR@:${RCDIR}:g' \
+   -e 's:@SBINDIR@:${SBINDIR}:g' \
-e 's: vpn : ng[0-9]*&:g' \
${DIST}/$@.in > $@
 .endfor
___
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: r305012 - in head/lib/libc: amd64/sys arm/sys i386/sys include mips/sys powerpc/sys powerpc64/sys sparc64/sys sys

2016-08-31 Thread Bryan Drewery 
On 8/29/16 11:47 AM, Konstantin Belousov wrote:
> Author: kib
> Date: Mon Aug 29 18:47:51 2016
> New Revision: 305012
> URL: https://svnweb.freebsd.org/changeset/base/305012
> 
> Log:
>   Rewrite ptrace(2) wrappers in C.
>   Besides removing hand-translation to assembler, this also adds missing
>   wrappers for arm64 and risc-v.
>   
>   Reviewed by:emaste, jhb
>   Sponsored by:   The FreeBSD Foundation
>   MFC after:  1 week
>   Differential revision:  https://reviews.freebsd.org/D7694
> 
> Added:
>   head/lib/libc/sys/ptrace.c   (contents, props changed)
> Deleted:
>   head/lib/libc/amd64/sys/ptrace.S
>   head/lib/libc/arm/sys/ptrace.S
>   head/lib/libc/i386/sys/ptrace.S
>   head/lib/libc/mips/sys/ptrace.S
>   head/lib/libc/powerpc/sys/ptrace.S
>   head/lib/libc/powerpc64/sys/ptrace.S
>   head/lib/libc/sparc64/sys/ptrace.S

There is a known bug with "FAST_DEPEND" (which is default and the only
way now) which may result in something like:

> cc: error: no such file or directory: '/usr/src/lib/libc/amd64/sys/ptrace.S'

I am working on a fix, but in the meantime a workaround is 'make
cleandepend', or removing the .depend.ptrace.* files in the lib/libc
object directories.



-- 
Regards,
Bryan Drewery



signature.asc
Description: OpenPGP digital signature

svn commit: r305164 - head/sys/kern

2016-08-31 Thread Mark Johnston 
Author: markj
Date: Wed Aug 31 21:48:22 2016
New Revision: 305164
URL: https://svnweb.freebsd.org/changeset/base/305164

Log:
  Rename unp_dispose_so() to unp_dispose().
  
  It implements the dom_dispose method for local socket domain, so its name
  should match the method name.

Modified:
  head/sys/kern/uipc_usrreq.c

Modified: head/sys/kern/uipc_usrreq.c
==
--- head/sys/kern/uipc_usrreq.c Wed Aug 31 21:41:32 2016(r305163)
+++ head/sys/kern/uipc_usrreq.c Wed Aug 31 21:48:22 2016(r305164)
@@ -277,8 +277,8 @@ static int  unp_connectat(int, struct soc
struct thread *);
 static int unp_connect2(struct socket *so, struct socket *so2, int);
 static voidunp_disconnect(struct unpcb *unp, struct unpcb *unp2);
-static voidunp_dispose(struct mbuf *);
-static voidunp_dispose_so(struct socket *so);
+static voidunp_dispose(struct socket *so);
+static voidunp_dispose_mbuf(struct mbuf *);
 static voidunp_shutdown(struct unpcb *);
 static voidunp_drop(struct unpcb *);
 static voidunp_gc(__unused void *, int);
@@ -335,7 +335,7 @@ static struct domain localdomain = {
.dom_name = "local",
.dom_init = unp_init,
.dom_externalize =  unp_externalize,
-   .dom_dispose =  unp_dispose_so,
+   .dom_dispose =  unp_dispose,
.dom_protosw =  localsw,
.dom_protoswNPROTOSW =  [nitems(localsw)]
 };
@@ -1051,7 +1051,7 @@ uipc_send(struct socket *so, int flags, 
UNP_LINK_RUNLOCK();
 
if (control != NULL && error != 0)
-   unp_dispose(control);
+   unp_dispose_mbuf(control);
 
 release:
if (control != NULL)
@@ -2352,7 +2352,7 @@ unp_gc(__unused void *arg, int pending)
 }
 
 static void
-unp_dispose(struct mbuf *m)
+unp_dispose_mbuf(struct mbuf *m)
 {
 
if (m)
@@ -2363,7 +2363,7 @@ unp_dispose(struct mbuf *m)
  * Synchronize against unp_gc, which can trip over data as we are freeing it.
  */
 static void
-unp_dispose_so(struct socket *so)
+unp_dispose(struct socket *so)
 {
struct unpcb *unp;
 
@@ -2371,7 +2371,7 @@ unp_dispose_so(struct socket *so)
UNP_LIST_LOCK();
unp->unp_gcflag |= UNPGC_IGNORE_RIGHTS;
UNP_LIST_UNLOCK();
-   unp_dispose(so->so_rcv.sb_mb);
+   unp_dispose_mbuf(so->so_rcv.sb_mb);
 }
 
 static void
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"

svn commit: r305163 - vendor-sys/acpica/20160831

2016-08-31 Thread Jung-uk Kim 
Author: jkim
Date: Wed Aug 31 21:41:32 2016
New Revision: 305163
URL: https://svnweb.freebsd.org/changeset/base/305163

Log:
  Tag ACPICA 20160831.

Added:
  vendor-sys/acpica/20160831/
 - copied from r305162, vendor-sys/acpica/dist/
___
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: r305162 - in vendor-sys/acpica/dist: . generate/unix/acpidump generate/unix/acpiexamples generate/unix/acpiexec generate/unix/acpisrc generate/unix/acpixtract generate/unix/iasl source/...

2016-08-31 Thread Jung-uk Kim 
Author: jkim
Date: Wed Aug 31 21:40:51 2016
New Revision: 305162
URL: https://svnweb.freebsd.org/changeset/base/305162

Log:
  Import ACPICA 20160831.

Added:
  vendor-sys/acpica/dist/source/compiler/aslhelp.c   (contents, props changed)
  vendor-sys/acpica/dist/source/components/utilities/utstrtoul64.c   (contents, 
props changed)
Modified:
  vendor-sys/acpica/dist/changes.txt
  vendor-sys/acpica/dist/generate/unix/acpidump/Makefile
  vendor-sys/acpica/dist/generate/unix/acpiexamples/Makefile
  vendor-sys/acpica/dist/generate/unix/acpiexec/Makefile
  vendor-sys/acpica/dist/generate/unix/acpisrc/Makefile
  vendor-sys/acpica/dist/generate/unix/acpixtract/Makefile
  vendor-sys/acpica/dist/generate/unix/iasl/Makefile
  vendor-sys/acpica/dist/source/compiler/aslcompiler.h
  vendor-sys/acpica/dist/source/compiler/aslmain.c
  vendor-sys/acpica/dist/source/compiler/aslopt.c
  vendor-sys/acpica/dist/source/compiler/asloptions.c
  vendor-sys/acpica/dist/source/compiler/aslstubs.c
  vendor-sys/acpica/dist/source/compiler/aslutils.c
  vendor-sys/acpica/dist/source/compiler/dtcompiler.h
  vendor-sys/acpica/dist/source/compiler/dtfield.c
  vendor-sys/acpica/dist/source/compiler/dtparser.y
  vendor-sys/acpica/dist/source/compiler/dtutils.c
  vendor-sys/acpica/dist/source/compiler/prparser.y
  vendor-sys/acpica/dist/source/components/debugger/dbconvert.c
  vendor-sys/acpica/dist/source/components/debugger/dbexec.c
  vendor-sys/acpica/dist/source/components/debugger/dbinput.c
  vendor-sys/acpica/dist/source/components/debugger/dbmethod.c
  vendor-sys/acpica/dist/source/components/disassembler/dmresrcl2.c
  vendor-sys/acpica/dist/source/components/dispatcher/dsmethod.c
  vendor-sys/acpica/dist/source/components/dispatcher/dswexec.c
  vendor-sys/acpica/dist/source/components/dispatcher/dswload2.c
  vendor-sys/acpica/dist/source/components/events/evgpe.c
  vendor-sys/acpica/dist/source/components/events/evgpeinit.c
  vendor-sys/acpica/dist/source/components/events/evrgnini.c
  vendor-sys/acpica/dist/source/components/events/evxfgpe.c
  vendor-sys/acpica/dist/source/components/executer/exconcat.c
  vendor-sys/acpica/dist/source/components/executer/exconfig.c
  vendor-sys/acpica/dist/source/components/executer/exconvrt.c
  vendor-sys/acpica/dist/source/components/executer/exmisc.c
  vendor-sys/acpica/dist/source/components/executer/exoparg1.c
  vendor-sys/acpica/dist/source/components/executer/exresop.c
  vendor-sys/acpica/dist/source/components/executer/extrace.c
  vendor-sys/acpica/dist/source/components/executer/exutils.c
  vendor-sys/acpica/dist/source/components/hardware/hwgpe.c
  vendor-sys/acpica/dist/source/components/namespace/nsconvert.c
  vendor-sys/acpica/dist/source/components/namespace/nsload.c
  vendor-sys/acpica/dist/source/components/namespace/nsparse.c
  vendor-sys/acpica/dist/source/components/namespace/nsutils.c
  vendor-sys/acpica/dist/source/components/parser/psparse.c
  vendor-sys/acpica/dist/source/components/parser/psxface.c
  vendor-sys/acpica/dist/source/components/tables/tbdata.c
  vendor-sys/acpica/dist/source/components/tables/tbfadt.c
  vendor-sys/acpica/dist/source/components/tables/tbfind.c
  vendor-sys/acpica/dist/source/components/tables/tbinstal.c
  vendor-sys/acpica/dist/source/components/tables/tbxfload.c
  vendor-sys/acpica/dist/source/components/utilities/utaddress.c
  vendor-sys/acpica/dist/source/components/utilities/uthex.c
  vendor-sys/acpica/dist/source/components/utilities/utnonansi.c
  vendor-sys/acpica/dist/source/components/utilities/utosi.c
  vendor-sys/acpica/dist/source/components/utilities/utxfinit.c
  vendor-sys/acpica/dist/source/include/acdebug.h
  vendor-sys/acpica/dist/source/include/acevents.h
  vendor-sys/acpica/dist/source/include/aclocal.h
  vendor-sys/acpica/dist/source/include/acnamesp.h
  vendor-sys/acpica/dist/source/include/acparser.h
  vendor-sys/acpica/dist/source/include/acpixf.h
  vendor-sys/acpica/dist/source/include/actables.h
  vendor-sys/acpica/dist/source/include/actypes.h
  vendor-sys/acpica/dist/source/include/acutils.h
  vendor-sys/acpica/dist/source/include/platform/acefi.h
  vendor-sys/acpica/dist/source/include/platform/acefiex.h
  vendor-sys/acpica/dist/source/tools/acpidump/apdump.c
  vendor-sys/acpica/dist/source/tools/acpidump/apmain.c
  vendor-sys/acpica/dist/source/tools/acpiexec/aeexec.c
  vendor-sys/acpica/dist/source/tools/acpiexec/aehandlers.c
  vendor-sys/acpica/dist/source/tools/acpiexec/aeinitfile.c
  vendor-sys/acpica/dist/source/tools/acpiexec/aemain.c

Modified: vendor-sys/acpica/dist/changes.txt
==
--- vendor-sys/acpica/dist/changes.txt  Wed Aug 31 21:35:51 2016
(r305161)
+++ vendor-sys/acpica/dist/changes.txt  Wed Aug 31 21:40:51 2016
(r305162)
@@ -1,7 +1,72 @@
 
-29 July 2016. Summary of changes for version 20160729:
+31 August 2016. Summary of changes for version 20160831:
 
-This release

svn commit: r305161 - in stable/10/sys: kern sys

2016-08-31 Thread Mark Johnston 
Author: markj
Date: Wed Aug 31 21:35:51 2016
New Revision: 305161
URL: https://svnweb.freebsd.org/changeset/base/305161

Log:
  MFC 303855:
  Handle races with listening socket close when connecting a unix socket.
  
  PR:   211531

Modified:
  stable/10/sys/kern/uipc_usrreq.c
  stable/10/sys/sys/unpcb.h
Directory Properties:
  stable/10/   (props changed)

Modified: stable/10/sys/kern/uipc_usrreq.c
==
--- stable/10/sys/kern/uipc_usrreq.cWed Aug 31 21:35:38 2016
(r305160)
+++ stable/10/sys/kern/uipc_usrreq.cWed Aug 31 21:35:51 2016
(r305161)
@@ -429,6 +429,8 @@ uipc_attach(struct socket *so, int proto
unp->unp_socket = so;
so->so_pcb = unp;
unp->unp_refcount = 1;
+   if (so->so_head != NULL)
+   unp->unp_flags |= UNP_NASCENT;
 
UNP_LIST_LOCK();
unp->unp_gencnt = ++unp_gencnt;
@@ -651,14 +653,22 @@ uipc_detach(struct socket *so)
unp = sotounpcb(so);
KASSERT(unp != NULL, ("uipc_detach: unp == NULL"));
 
-   UNP_LINK_WLOCK();
+   vp = NULL;
+   local_unp_rights = 0;
+
UNP_LIST_LOCK();
-   UNP_PCB_LOCK(unp);
LIST_REMOVE(unp, unp_link);
unp->unp_gencnt = ++unp_gencnt;
--unp_count;
UNP_LIST_UNLOCK();
 
+   if ((unp->unp_flags & UNP_NASCENT) != 0) {
+   UNP_PCB_LOCK(unp);
+   goto teardown;
+   }
+   UNP_LINK_WLOCK();
+   UNP_PCB_LOCK(unp);
+
/*
 * XXXRW: Should assert vp->v_socket == so.
 */
@@ -686,6 +696,7 @@ uipc_detach(struct socket *so)
}
local_unp_rights = unp_rights;
UNP_LINK_WUNLOCK();
+teardown:
unp->unp_socket->so_pcb = NULL;
saved_unp_addr = unp->unp_addr;
unp->unp_addr = NULL;
@@ -1441,6 +1452,7 @@ unp_connect2(struct socket *so, struct s
 
if (so2->so_type != so->so_type)
return (EPROTOTYPE);
+   unp2->unp_flags &= ~UNP_NASCENT;
unp->unp_conn = unp2;
 
switch (so->so_type) {

Modified: stable/10/sys/sys/unpcb.h
==
--- stable/10/sys/sys/unpcb.h   Wed Aug 31 21:35:38 2016(r305160)
+++ stable/10/sys/sys/unpcb.h   Wed Aug 31 21:35:51 2016(r305161)
@@ -103,10 +103,6 @@ struct unpcb {
 #defineUNP_WANTCRED0x004   /* credentials wanted */
 #defineUNP_CONNWAIT0x008   /* connect blocks until 
accepted */
 
-#defineUNPGC_REF   0x1 /* unpcb has external 
ref. */
-#defineUNPGC_DEAD  0x2 /* unpcb might be dead. 
*/
-#defineUNPGC_SCANNED   0x4 /* Has been scanned. */
-
 /*
  * These flags are used to handle non-atomicity in connect() and bind()
  * operations on a socket: in particular, to avoid races between multiple
@@ -114,6 +110,14 @@ struct unpcb {
  */
 #defineUNP_CONNECTING  0x010   /* Currently 
connecting. */
 #defineUNP_BINDING 0x020   /* Currently binding. */
+#defineUNP_NASCENT 0x040   /* Newborn child 
socket. */
+
+/*
+ * Flags in unp_gcflag.
+ */
+#defineUNPGC_REF   0x1 /* unpcb has external 
ref. */
+#defineUNPGC_DEAD  0x2 /* unpcb might be dead. 
*/
+#defineUNPGC_SCANNED   0x4 /* Has been scanned. */
 
 #definesotounpcb(so)   ((struct unpcb *)((so)->so_pcb))
 
___
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: r305160 - in head/sys/boot/efi: boot1 loader

2016-08-31 Thread Ed Maste 
Author: emaste
Date: Wed Aug 31 21:35:38 2016
New Revision: 305160
URL: https://svnweb.freebsd.org/changeset/base/305160

Log:
  Set UEFI boot loader PE/COFF timestamps to known value for reproducible builds
  
  Reviewed by:  rpokala
  MFC after:1 week
  Sponsored by: The FreeBSD Foundation
  Differential Revision:https://reviews.freebsd.org/D7633

Modified:
  head/sys/boot/efi/boot1/Makefile
  head/sys/boot/efi/loader/Makefile

Modified: head/sys/boot/efi/boot1/Makefile
==
--- head/sys/boot/efi/boot1/MakefileWed Aug 31 21:35:12 2016
(r305159)
+++ head/sys/boot/efi/boot1/MakefileWed Aug 31 21:35:38 2016
(r305160)
@@ -89,11 +89,15 @@ EFI_TARGET= efi-app-ia32
 EFI_TARGET=binary
 .endif
 
+# Arbitrarily set the PE/COFF header timestamps to 1 Jan 2016 00:00:00
+# for build reproducibility.
+SOURCE_DATE_EPOCH?=1451606400
 boot1.efi: ${PROG}
if ${NM} ${.ALLSRC} | grep ' U '; then \
echo "Undefined symbols in ${.ALLSRC}"; \
exit 1; \
fi
+   SOURCE_DATE_EPOCH=${SOURCE_DATE_EPOCH} \
${OBJCOPY} -j .peheader -j .text -j .sdata -j .data \
-j .dynamic -j .dynsym -j .rel.dyn \
-j .rela.dyn -j .reloc -j .eh_frame \

Modified: head/sys/boot/efi/loader/Makefile
==
--- head/sys/boot/efi/loader/Makefile   Wed Aug 31 21:35:12 2016
(r305159)
+++ head/sys/boot/efi/loader/Makefile   Wed Aug 31 21:35:38 2016
(r305160)
@@ -131,11 +131,15 @@ EFI_TARGET=   efi-app-ia32
 EFI_TARGET=binary
 .endif
 
+# Arbitrarily set the PE/COFF header timestamps to 1 Jan 2016 00:00:00
+# for build reproducibility.
+SOURCE_DATE_EPOCH?=1451606400
 loader.efi: ${PROG}
if ${NM} ${.ALLSRC} | grep ' U '; then \
echo "Undefined symbols in ${.ALLSRC}"; \
exit 1; \
fi
+   SOURCE_DATE_EPOCH=${SOURCE_DATE_EPOCH} \
${OBJCOPY} -j .peheader -j .text -j .sdata -j .data \
-j .dynamic -j .dynsym -j .rel.dyn \
-j .rela.dyn -j .reloc -j .eh_frame -j set_Xcommand_set \
___
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: r305159 - in stable/11/sys: kern sys

2016-08-31 Thread Mark Johnston 
Author: markj
Date: Wed Aug 31 21:35:12 2016
New Revision: 305159
URL: https://svnweb.freebsd.org/changeset/base/305159

Log:
  MFC r303855:
  Handle races with listening socket close when connecting a unix socket.
  
  PR:   211531

Modified:
  stable/11/sys/kern/uipc_usrreq.c
  stable/11/sys/sys/unpcb.h
Directory Properties:
  stable/11/   (props changed)

Modified: stable/11/sys/kern/uipc_usrreq.c
==
--- stable/11/sys/kern/uipc_usrreq.cWed Aug 31 21:18:38 2016
(r305158)
+++ stable/11/sys/kern/uipc_usrreq.cWed Aug 31 21:35:12 2016
(r305159)
@@ -430,6 +430,8 @@ uipc_attach(struct socket *so, int proto
unp->unp_socket = so;
so->so_pcb = unp;
unp->unp_refcount = 1;
+   if (so->so_head != NULL)
+   unp->unp_flags |= UNP_NASCENT;
 
UNP_LIST_LOCK();
unp->unp_gencnt = ++unp_gencnt;
@@ -652,14 +654,22 @@ uipc_detach(struct socket *so)
unp = sotounpcb(so);
KASSERT(unp != NULL, ("uipc_detach: unp == NULL"));
 
-   UNP_LINK_WLOCK();
+   vp = NULL;
+   local_unp_rights = 0;
+
UNP_LIST_LOCK();
-   UNP_PCB_LOCK(unp);
LIST_REMOVE(unp, unp_link);
unp->unp_gencnt = ++unp_gencnt;
--unp_count;
UNP_LIST_UNLOCK();
 
+   if ((unp->unp_flags & UNP_NASCENT) != 0) {
+   UNP_PCB_LOCK(unp);
+   goto teardown;
+   }
+   UNP_LINK_WLOCK();
+   UNP_PCB_LOCK(unp);
+
/*
 * XXXRW: Should assert vp->v_socket == so.
 */
@@ -687,6 +697,7 @@ uipc_detach(struct socket *so)
}
local_unp_rights = unp_rights;
UNP_LINK_WUNLOCK();
+teardown:
unp->unp_socket->so_pcb = NULL;
saved_unp_addr = unp->unp_addr;
unp->unp_addr = NULL;
@@ -1473,6 +1484,7 @@ unp_connect2(struct socket *so, struct s
 
if (so2->so_type != so->so_type)
return (EPROTOTYPE);
+   unp2->unp_flags &= ~UNP_NASCENT;
unp->unp_conn = unp2;
 
switch (so->so_type) {

Modified: stable/11/sys/sys/unpcb.h
==
--- stable/11/sys/sys/unpcb.h   Wed Aug 31 21:18:38 2016(r305158)
+++ stable/11/sys/sys/unpcb.h   Wed Aug 31 21:35:12 2016(r305159)
@@ -103,11 +103,6 @@ struct unpcb {
 #defineUNP_WANTCRED0x004   /* credentials wanted */
 #defineUNP_CONNWAIT0x008   /* connect blocks until 
accepted */
 
-#defineUNPGC_REF   0x1 /* unpcb has external 
ref. */
-#defineUNPGC_DEAD  0x2 /* unpcb might be dead. 
*/
-#defineUNPGC_SCANNED   0x4 /* Has been scanned. */
-#defineUNPGC_IGNORE_RIGHTS 0x8 /* Attached rights are 
freed */
-
 /*
  * These flags are used to handle non-atomicity in connect() and bind()
  * operations on a socket: in particular, to avoid races between multiple
@@ -115,6 +110,15 @@ struct unpcb {
  */
 #defineUNP_CONNECTING  0x010   /* Currently 
connecting. */
 #defineUNP_BINDING 0x020   /* Currently binding. */
+#defineUNP_NASCENT 0x040   /* Newborn child 
socket. */
+
+/*
+ * Flags in unp_gcflag.
+ */
+#defineUNPGC_REF   0x1 /* unpcb has external 
ref. */
+#defineUNPGC_DEAD  0x2 /* unpcb might be dead. 
*/
+#defineUNPGC_SCANNED   0x4 /* Has been scanned. */
+#defineUNPGC_IGNORE_RIGHTS 0x8 /* Attached rights are 
freed */
 
 #definesotounpcb(so)   ((struct unpcb *)((so)->so_pcb))
 
___
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: r304510 - head/sys/dev/ixl

2016-08-31 Thread Dimitry Andric 
On 31 Aug 2016, at 18:53, Ed Maste  wrote:
> 
> On 19 August 2016 at 20:08, Eric Joyner  wrote:
>> Author: erj
>> Date: Sat Aug 20 00:08:10 2016
>> New Revision: 304510
>> URL: https://svnweb.freebsd.org/changeset/base/304510
>> 
>> Log:
>>  ixlv(4): Fix ixlv(4) not loading when loaded as a kernel module and netmap 
>> is enabled.
> 
> I did not confirm it's this commit, but kernel builds currently have
> this warning:
> 
> if_ixlv.o: warning: common of `ixl_crcstrip' overridden by definition
> if_ixl.o: warning: defined here
> if_ixlv.o: warning: multiple common of `ixl_rx_miss'
> if_ixl.o: warning: previous common is here
> if_ixlv.o: warning: multiple common of `ixl_rx_miss_bufs'
> if_ixl.o: warning: previous common is here

Interestingly, with the external gcc build, it is even an error:

--- kernel.full ---
linking kernel.full
if_ixlv.o:(.bss+0x0): multiple definition of `ixl_crcstrip'
if_ixl.o:(.data+0xb8): first defined here
if_ixlv.o: In function `ixlv_media_change':
/builds/FreeBSD_HEAD_amd64_gcc/sys/dev/ixl/if_ixlv.c:2252: multiple definition 
of `ixl_rx_miss_bufs'
if_ixl.o:/builds/FreeBSD_HEAD_amd64_gcc/sys/dev/ixl/if_ixl.c:680: first defined 
here
if_ixlv.o: In function `ixlv_media_change':
/builds/FreeBSD_HEAD_amd64_gcc/sys/dev/ixl/if_ixlv.c:2262: multiple definition 
of `ixl_rx_miss'
if_ixl.o:/builds/FreeBSD_HEAD_amd64_gcc/sys/dev/ixl/if_ixl.c:681: first defined 
here
*** [kernel.full] Error code 1

See e.g. https://jenkins.freebsd.org/job/FreeBSD_HEAD_amd64_gcc/1505/console

-Dimitry



signature.asc
Description: Message signed with OpenPGP using GPGMail

svn commit: r305157 - stable/11/sys/vm

2016-08-31 Thread Mark Johnston 
Author: markj
Date: Wed Aug 31 21:14:16 2016
New Revision: 305157
URL: https://svnweb.freebsd.org/changeset/base/305157

Log:
  MFC r304053, r304054:
  Initialize busy lock state and strengthen busy lock assertions.

Modified:
  stable/11/sys/vm/vm_page.c
  stable/11/sys/vm/vm_phys.c
Directory Properties:
  stable/11/   (props changed)

Modified: stable/11/sys/vm/vm_page.c
==
--- stable/11/sys/vm/vm_page.c  Wed Aug 31 21:12:50 2016(r305156)
+++ stable/11/sys/vm/vm_page.c  Wed Aug 31 21:14:16 2016(r305157)
@@ -1712,8 +1712,7 @@ vm_page_alloc(vm_object_t object, vm_pin
("vm_page_alloc: page %p has unexpected queue %d", m, m->queue));
KASSERT(m->wire_count == 0, ("vm_page_alloc: page %p is wired", m));
KASSERT(m->hold_count == 0, ("vm_page_alloc: page %p is held", m));
-   KASSERT(!vm_page_sbusied(m),
-   ("vm_page_alloc: page %p is busy", m));
+   KASSERT(!vm_page_busied(m), ("vm_page_alloc: page %p is busy", m));
KASSERT(m->dirty == 0, ("vm_page_alloc: page %p is dirty", m));
KASSERT(pmap_page_get_memattr(m) == VM_MEMATTR_DEFAULT,
("vm_page_alloc: page %p has unexpected memattr %d", m,
@@ -2021,7 +2020,7 @@ vm_page_alloc_init(vm_page_t m)
("vm_page_alloc_init: page %p is wired", m));
KASSERT(m->hold_count == 0,
("vm_page_alloc_init: page %p is held", m));
-   KASSERT(!vm_page_sbusied(m),
+   KASSERT(!vm_page_busied(m),
("vm_page_alloc_init: page %p is busy", m));
KASSERT(m->dirty == 0,
("vm_page_alloc_init: page %p is dirty", m));

Modified: stable/11/sys/vm/vm_phys.c
==
--- stable/11/sys/vm/vm_phys.c  Wed Aug 31 21:12:50 2016(r305156)
+++ stable/11/sys/vm/vm_phys.c  Wed Aug 31 21:14:16 2016(r305157)
@@ -741,6 +741,7 @@ vm_phys_add_page(vm_paddr_t pa)
 
vm_cnt.v_page_count++;
m = vm_phys_paddr_to_vm_page(pa);
+   m->busy_lock = VPB_UNBUSIED;
m->phys_addr = pa;
m->queue = PQ_NONE;
m->segind = vm_phys_paddr_to_segind(pa);
___
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: r305156 - head/usr.bin

2016-08-31 Thread Ed Maste 
Author: emaste
Date: Wed Aug 31 21:12:50 2016
New Revision: 305156
URL: https://svnweb.freebsd.org/changeset/base/305156

Log:
  users: don't build when we don't have a C++ toolchain
  
  (On RISC-V MK_CXX is in BROKEN_OPTIONS, so users remains skipped there.)
  
  Reviewed by:  bdrewery
  Differential Revision:https://reviews.freebsd.org/D7725

Modified:
  head/usr.bin/Makefile

Modified: head/usr.bin/Makefile
==
--- head/usr.bin/Makefile   Wed Aug 31 21:04:58 2016(r305155)
+++ head/usr.bin/Makefile   Wed Aug 31 21:12:50 2016(r305156)
@@ -294,7 +294,7 @@ SUBDIR.${MK_VT}+=   vtfontcvt
 SUBDIR.${MK_USB}+= usbhidaction
 SUBDIR.${MK_USB}+= usbhidctl
 SUBDIR.${MK_UTMPX}+=   last
-.if ${MACHINE_CPUARCH} != "riscv" # RISCVTODO users does not build
+.if ${MK_CXX} != "no"
 SUBDIR.${MK_UTMPX}+=   users
 .endif
 SUBDIR.${MK_UTMPX}+=   who
___
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: r305155 - head/contrib/elftoolchain/readelf

2016-08-31 Thread Ed Maste 
Author: emaste
Date: Wed Aug 31 21:04:58 2016
New Revision: 305155
URL: https://svnweb.freebsd.org/changeset/base/305155

Log:
  readelf: silence GCC 4.2.1 uninitialized variable warning
  
  Sponsored by: The FreeBSD Foundation

Modified:
  head/contrib/elftoolchain/readelf/readelf.c

Modified: head/contrib/elftoolchain/readelf/readelf.c
==
--- head/contrib/elftoolchain/readelf/readelf.c Wed Aug 31 20:38:40 2016
(r305154)
+++ head/contrib/elftoolchain/readelf/readelf.c Wed Aug 31 21:04:58 2016
(r305155)
@@ -2831,6 +2831,8 @@ dump_rel(struct readelf *re, struct sect
type2 = (type >> 8) & 0xFF;
type3 = (type >> 16) & 0xFF;
type = type & 0xFF;
+   } else {
+   type2 = type3 = 0;
}
if (re->options & RE_WW)
printf("%16.16jx %16.16jx %-24.24s"
@@ -2914,6 +2916,8 @@ dump_rela(struct readelf *re, struct sec
type2 = (type >> 8) & 0xFF;
type3 = (type >> 16) & 0xFF;
type = type & 0xFF;
+   } else {
+   type2 = type3 = 0;
}
if (re->options & RE_WW)
printf("%16.16jx %16.16jx %-24.24s"
___
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: r305154 - head/lib/libc/gen

2016-08-31 Thread Jilles Tjoelker 
Author: jilles
Date: Wed Aug 31 20:38:40 2016
New Revision: 305154
URL: https://svnweb.freebsd.org/changeset/base/305154

Log:
  directory(3): Deprecate readdir_r(). Clarify dirent buffers.
  
  In existing implementations including FreeBSD, there is no reason to use
  readdir_r() in the common case where potentially multiple threads each list
  their own directory. Code using readdir() is simpler.
  
  What's more, lthough readdir_r() can safely be used on FreeBSD because
  NAME_MAX is forced to 255, it cannot be used safely on systems where
  {NAME_MAX} is not fixed. As a concrete example, FAT/NTFS filenames can be up
  to 255 UTF-16 code units long, which can be up to 765 UTF-8 bytes.
  
  Deprecating readdir_r() in POSIX has been proposed in
  http://www.austingroupbugs.net/view.php?id=696
  and glibc wants to deprecate it as well.
  
  Reviewed by:  ed, wblock
  MFC after:1 week
  Differential Revision:https://reviews.freebsd.org/D7678

Modified:
  head/lib/libc/gen/directory.3

Modified: head/lib/libc/gen/directory.3
==
--- head/lib/libc/gen/directory.3   Wed Aug 31 20:33:59 2016
(r305153)
+++ head/lib/libc/gen/directory.3   Wed Aug 31 20:38:40 2016
(r305154)
@@ -28,7 +28,7 @@
 .\" @(#)directory.38.1 (Berkeley) 6/4/93
 .\" $FreeBSD$
 .\"
-.Dd February 19, 2016
+.Dd August 31, 2016
 .Dt DIRECTORY 3
 .Os
 .Sh NAME
@@ -68,6 +68,15 @@
 .Ft int
 .Fn dirfd "DIR *dirp"
 .Sh DESCRIPTION
+.Bf -symbolic
+The
+.Fn readdir_r
+interface is deprecated
+because it cannot be used correctly unless
+.Brq Va NAME_MAX
+is a fixed value.
+.Ef
+.Pp
 The
 .Fn opendir
 function
@@ -122,7 +131,13 @@ The
 .Fn readdir
 function
 returns a pointer to the next directory entry.
-It returns
+The directory entry remains valid until the next call to
+.Fn readdir
+or
+.Fn closedir
+on the same
+.Em directory stream .
+The function returns
 .Dv NULL
 upon reaching the end of the directory or on error.
 In the event of an error,
@@ -139,6 +154,13 @@ provides the same functionality as
 but the caller must provide a directory
 .Fa entry
 buffer to store the results in.
+The buffer must be large enough for a
+.Vt struct dirent
+with a
+.Va d_name
+array with
+.Brq Va NAME_MAX
++ 1 elements.
 If the read succeeds,
 .Fa result
 is pointed at 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: r305153 - in stable/11/secure/lib/libcrypto: . amd64 arm i386

2016-08-31 Thread Jung-uk Kim 
Author: jkim
Date: Wed Aug 31 20:33:59 2016
New Revision: 305153
URL: https://svnweb.freebsd.org/changeset/base/305153

Log:
  MFC:  r304638, r304640
  
  Fix white spaces and prefer C-style comments in assembly sources.

Modified:
  stable/11/secure/lib/libcrypto/Makefile.asm
  stable/11/secure/lib/libcrypto/amd64/aes-x86_64.S
  stable/11/secure/lib/libcrypto/amd64/aesni-gcm-x86_64.S
  stable/11/secure/lib/libcrypto/amd64/aesni-mb-x86_64.S
  stable/11/secure/lib/libcrypto/amd64/aesni-sha1-x86_64.S
  stable/11/secure/lib/libcrypto/amd64/aesni-sha256-x86_64.S
  stable/11/secure/lib/libcrypto/amd64/aesni-x86_64.S
  stable/11/secure/lib/libcrypto/amd64/bsaes-x86_64.S
  stable/11/secure/lib/libcrypto/amd64/cmll-x86_64.S
  stable/11/secure/lib/libcrypto/amd64/ecp_nistz256-x86_64.S
  stable/11/secure/lib/libcrypto/amd64/ghash-x86_64.S
  stable/11/secure/lib/libcrypto/amd64/md5-x86_64.S
  stable/11/secure/lib/libcrypto/amd64/rc4-md5-x86_64.S
  stable/11/secure/lib/libcrypto/amd64/rc4-x86_64.S
  stable/11/secure/lib/libcrypto/amd64/rsaz-avx2.S
  stable/11/secure/lib/libcrypto/amd64/rsaz-x86_64.S
  stable/11/secure/lib/libcrypto/amd64/sha1-mb-x86_64.S
  stable/11/secure/lib/libcrypto/amd64/sha1-x86_64.S
  stable/11/secure/lib/libcrypto/amd64/sha256-mb-x86_64.S
  stable/11/secure/lib/libcrypto/amd64/sha256-x86_64.S
  stable/11/secure/lib/libcrypto/amd64/sha512-x86_64.S
  stable/11/secure/lib/libcrypto/amd64/vpaes-x86_64.S
  stable/11/secure/lib/libcrypto/amd64/wp-x86_64.S
  stable/11/secure/lib/libcrypto/amd64/x86_64-gf2m.S
  stable/11/secure/lib/libcrypto/amd64/x86_64-mont.S
  stable/11/secure/lib/libcrypto/amd64/x86_64-mont5.S
  stable/11/secure/lib/libcrypto/amd64/x86_64cpuid.S
  stable/11/secure/lib/libcrypto/arm/aes-armv4.S
  stable/11/secure/lib/libcrypto/arm/aesv8-armx.S
  stable/11/secure/lib/libcrypto/arm/armv4-gf2m.S
  stable/11/secure/lib/libcrypto/arm/armv4-mont.S
  stable/11/secure/lib/libcrypto/arm/bsaes-armv7.S
  stable/11/secure/lib/libcrypto/arm/ghash-armv4.S
  stable/11/secure/lib/libcrypto/arm/ghashv8-armx.S
  stable/11/secure/lib/libcrypto/arm/sha1-armv4-large.S
  stable/11/secure/lib/libcrypto/arm/sha256-armv4.S
  stable/11/secure/lib/libcrypto/arm/sha512-armv4.S
  stable/11/secure/lib/libcrypto/i386/aes-586.S
  stable/11/secure/lib/libcrypto/i386/aesni-x86.S
  stable/11/secure/lib/libcrypto/i386/bf-586.S
  stable/11/secure/lib/libcrypto/i386/bf-686.S
  stable/11/secure/lib/libcrypto/i386/bn-586.S
  stable/11/secure/lib/libcrypto/i386/cmll-x86.S
  stable/11/secure/lib/libcrypto/i386/co-586.S
  stable/11/secure/lib/libcrypto/i386/crypt586.S
  stable/11/secure/lib/libcrypto/i386/des-586.S
  stable/11/secure/lib/libcrypto/i386/ghash-x86.S
  stable/11/secure/lib/libcrypto/i386/md5-586.S
  stable/11/secure/lib/libcrypto/i386/rc4-586.S
  stable/11/secure/lib/libcrypto/i386/rc5-586.S
  stable/11/secure/lib/libcrypto/i386/rmd-586.S
  stable/11/secure/lib/libcrypto/i386/sha1-586.S
  stable/11/secure/lib/libcrypto/i386/sha256-586.S
  stable/11/secure/lib/libcrypto/i386/sha512-586.S
  stable/11/secure/lib/libcrypto/i386/vpaes-x86.S
  stable/11/secure/lib/libcrypto/i386/wp-mmx.S
  stable/11/secure/lib/libcrypto/i386/x86-gf2m.S
  stable/11/secure/lib/libcrypto/i386/x86-mont.S
  stable/11/secure/lib/libcrypto/i386/x86cpuid.S
Directory Properties:
  stable/11/   (props changed)

Modified: stable/11/secure/lib/libcrypto/Makefile.asm
==
--- stable/11/secure/lib/libcrypto/Makefile.asm Wed Aug 31 20:30:49 2016
(r305152)
+++ stable/11/secure/lib/libcrypto/Makefile.asm Wed Aug 31 20:33:59 2016
(r305153)
@@ -64,8 +64,8 @@ CLEANFILES=   ${ASM} ${SHA_ASM:S/$/.s/}
 .SUFFIXES: .pl
 
 .pl.S:
-   ( echo '# $$'FreeBSD'$$' ;\
-   echo '# Do not modify. This file is auto-generated from ${.IMPSRC:T}.' 
;\
+   ( echo '/* $$'FreeBSD'$$ */' ;\
+   echo '/* Do not modify. This file is auto-generated from ${.IMPSRC:T}. 
*/' ;\
env CC=cc perl ${.IMPSRC} elf ) > ${.TARGET}
 
 ${SHA_TMP}: ${SHA_SRC}
@@ -73,8 +73,8 @@ ${SHA_TMP}: ${SHA_SRC}
 
 .for s in ${SHA_ASM}
 ${s}.S: ${s}.s
-   ( echo '# $$'FreeBSD'$$' ;\
-   echo '  # Do not modify. This file is auto-generated from ${SHA_SRC}.' 
;\
+   ( echo '/* $$'FreeBSD'$$ */' ;\
+   echo '/* Do not modify. This file is auto-generated from ${SHA_SRC}. 
*/' ;\
cat ${s}.s ) > ${.TARGET}
 .endfor
 
@@ -108,14 +108,14 @@ CLEANFILES=   ${ASM} ${SRCS:R:S/$/.s/}
 .SUFFIXES: .pl
 
 aes-armv4.S:   aes-armv4.pl
-   ( echo '# $$'FreeBSD'$$' ;\
-   echo '# Do not modify. This file is auto-generated from ${.ALLSRC:T}.' 
;\
+   ( echo '/* $$'FreeBSD'$$ */' ;\
+   echo '/* Do not modify. This file is auto-generated from ${.ALLSRC:T}. 
*/' ;\
env CC=cc perl ${.ALLSRC} elf ) > ${.TARGET}
 
 .pl.S:
env CC=cc perl ${.IMPSRC} elf ${.TARGET:R:S/$/.s/}
-   ( echo '# $$'FreeBSD'$$' ;\
-   echo '  # Do not

svn commit: r305152 - in stable/11: crypto/openssl/crypto/bn/asm crypto/openssl/crypto/sha/asm secure/lib/libcrypto secure/lib/libcrypto/arm

2016-08-31 Thread Jung-uk Kim 
Author: jkim
Date: Wed Aug 31 20:30:49 2016
New Revision: 305152
URL: https://svnweb.freebsd.org/changeset/base/305152

Log:
  MFC:  r304636
  
  Build OpenSSL assembly sources for arm.

Added:
  stable/11/secure/lib/libcrypto/arm/
 - copied from r304636, head/secure/lib/libcrypto/arm/
Modified:
  stable/11/crypto/openssl/crypto/bn/asm/armv4-gf2m.pl
  stable/11/crypto/openssl/crypto/sha/asm/sha256-armv4.pl
  stable/11/secure/lib/libcrypto/Makefile
  stable/11/secure/lib/libcrypto/Makefile.asm
  stable/11/secure/lib/libcrypto/Makefile.inc
Directory Properties:
  stable/11/   (props changed)

Modified: stable/11/crypto/openssl/crypto/bn/asm/armv4-gf2m.pl
==
--- stable/11/crypto/openssl/crypto/bn/asm/armv4-gf2m.plWed Aug 31 
20:25:37 2016(r305151)
+++ stable/11/crypto/openssl/crypto/bn/asm/armv4-gf2m.plWed Aug 31 
20:30:49 2016(r305152)
@@ -213,8 +213,8 @@ $code.=<<___;
 .align 5
 .LNEON:
ldr r12, [sp]   @ 5th argument
-   vmov.32 $a, r2, r1
-   vmov.32 $b, r12, r3
+   vmov$a, r2, r1
+   vmov$b, r12, r3
vmov.i64$k48, #0x
vmov.i64$k32, #0x
vmov.i64$k16, #0x

Modified: stable/11/crypto/openssl/crypto/sha/asm/sha256-armv4.pl
==
--- stable/11/crypto/openssl/crypto/sha/asm/sha256-armv4.pl Wed Aug 31 
20:25:37 2016(r305151)
+++ stable/11/crypto/openssl/crypto/sha/asm/sha256-armv4.pl Wed Aug 31 
20:30:49 2016(r305152)
@@ -595,7 +595,7 @@ sha256_block_data_order_armv8:
adr $Ktbl,.LARMv8
sub $Ktbl,$Ktbl,#.LARMv8-K256
 # else
-   adrl$Ktbl,K256
+   sub $Ktbl,$Ktbl,#256+32
 # endif
add $len,$inp,$len,lsl#6@ len to point at the end of inp
 

Modified: stable/11/secure/lib/libcrypto/Makefile
==
--- stable/11/secure/lib/libcrypto/Makefile Wed Aug 31 20:25:37 2016
(r305151)
+++ stable/11/secure/lib/libcrypto/Makefile Wed Aug 31 20:30:49 2016
(r305152)
@@ -24,6 +24,8 @@ SRCS= cpt_err.c cryptlib.c cversion.c ex
o_fips.c o_init.c o_str.c o_time.c uid.c
 .if defined(ASM_amd64)
 SRCS+= x86_64cpuid.S
+.elif defined(ASM_arm)
+SRCS+= armcap.c armv4cpuid.S
 .elif defined(ASM_i386)
 SRCS+= x86cpuid.S
 .else
@@ -36,6 +38,8 @@ SRCS+=aes_cfb.c aes_ctr.c aes_ecb.c aes
 .if defined(ASM_amd64)
 SRCS+= aes-x86_64.S aesni-mb-x86_64.S aesni-sha1-x86_64.S \
aesni-sha256-x86_64.S aesni-x86_64.S bsaes-x86_64.S vpaes-x86_64.S
+.elif defined(ASM_arm)
+SRCS+= aes-armv4.S aes_cbc.c aesv8-armx.S bsaes-armv7.S
 .elif defined(ASM_i386)
 SRCS+= aes-586.S aesni-x86.S vpaes-x86.S
 .else
@@ -85,6 +89,8 @@ SRCS+=bn_add.c bn_blind.c bn_const.c bn
 .if defined(ASM_amd64)
 SRCS+= rsaz-avx2.S rsaz-x86_64.S rsaz_exp.c x86_64-gcc.c x86_64-gf2m.S \
x86_64-mont.S x86_64-mont5.S
+.elif defined(ASM_arm)
+SRCS+= armv4-mont.S armv4-gf2m.S bn_asm.c
 .elif defined(ASM_i386)
 SRCS+= bn-586.S co-586.S x86-gf2m.S x86-mont.S
 .else
@@ -234,6 +240,8 @@ SRCS+=  cbc128.c ccm128.c cfb128.c ctr128
wrap128.c xts128.c
 .if defined(ASM_amd64)
 SRCS+= aesni-gcm-x86_64.S ghash-x86_64.S
+.elif defined(ASM_arm)
+SRCS+= ghash-armv4.S ghashv8-armx.S
 .elif defined(ASM_i386)
 SRCS+= ghash-x86.S
 .endif
@@ -319,6 +327,8 @@ SRCS+=  sha1_one.c sha1dgst.c sha256.c sh
 .if defined(ASM_amd64)
 SRCS+= sha1-mb-x86_64.S sha1-x86_64.S sha256-mb-x86_64.S sha256-x86_64.S \
sha512-x86_64.S
+.elif defined(ASM_arm)
+SRCS+= sha1-armv4-large.S sha256-armv4.S sha512-armv4.S
 .elif defined(ASM_i386)
 SRCS+= sha1-586.S sha256-586.S sha512-586.S
 .endif

Modified: stable/11/secure/lib/libcrypto/Makefile.asm
==
--- stable/11/secure/lib/libcrypto/Makefile.asm Wed Aug 31 20:25:37 2016
(r305151)
+++ stable/11/secure/lib/libcrypto/Makefile.asm Wed Aug 31 20:30:49 2016
(r305152)
@@ -78,6 +78,46 @@ ${s}.S: ${s}.s
cat ${s}.s ) > ${.TARGET}
 .endfor
 
+.elif defined(ASM_arm)
+
+.PATH: ${LCRYPTO_SRC}/crypto \
+   ${LCRYPTO_SRC}/crypto/aes/asm \
+   ${LCRYPTO_SRC}/crypto/bn/asm \
+   ${LCRYPTO_SRC}/crypto/modes/asm \
+   ${LCRYPTO_SRC}/crypto/sha/asm
+
+PERLPATH=  -I${LCRYPTO_SRC}/crypto/perlasm
+
+# aes
+SRCS=  aesv8-armx.pl bsaes-armv7.pl
+
+# bn
+SRCS+= armv4-mont.pl armv4-gf2m.pl
+
+# modes
+SRCS+= ghash-armv4.pl ghashv8-armx.pl
+
+# sha
+SRCS+= sha1-armv4-large.pl sha256-armv4.pl sha512-armv4.pl
+
+ASM=   aes-armv4.S ${SRCS:R:S/$/.S/}
+
+all:   ${ASM}
+
+CLEANFILES=${ASM} ${SRCS:R:S/$/.s/}
+.SUFFIXES: .pl
+
+aes-armv4.S:   aes-armv4.pl
+   ( echo '# $$'FreeBSD'$$' ;\
+   echo

svn commit: r305151 - stable/11/share/man/man4

2016-08-31 Thread Konstantin Belousov 
Author: kib
Date: Wed Aug 31 20:25:37 2016
New Revision: 305151
URL: https://svnweb.freebsd.org/changeset/base/305151

Log:
  MFC r304315 (by jilles):
  rights(4): CAP_FSYNC also permits fdatasync(2).

Modified:
  stable/11/share/man/man4/rights.4
Directory Properties:
  stable/11/   (props changed)

Modified: stable/11/share/man/man4/rights.4
==
--- stable/11/share/man/man4/rights.4   Wed Aug 31 20:21:03 2016
(r305150)
+++ stable/11/share/man/man4/rights.4   Wed Aug 31 20:25:37 2016
(r305151)
@@ -32,7 +32,7 @@
 .\"
 .\" $FreeBSD$
 .\"
-.Dd August 27, 2015
+.Dd August 17, 2016
 .Dt RIGHTS 4
 .Os
 .Sh NAME
@@ -254,6 +254,7 @@ Permit
 .It Dv CAP_FSYNC
 Permit
 .Xr aio_fsync 2 ,
+.Xr fdatasync 2 ,
 .Xr fsync 2
 and
 .Xr openat 2
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"

svn commit: r305150 - stable/11/secure/lib/libcrypto

2016-08-31 Thread Jung-uk Kim 
Author: jkim
Date: Wed Aug 31 20:21:03 2016
New Revision: 305150
URL: https://svnweb.freebsd.org/changeset/base/305150

Log:
  MFC:  r304320
  
  Disable assembly sources when compiler/assembler cannot compile certain
  instructions.  For example, GCC 4.2.1 + binutils 2.17.50 does not support
  AVX instructions.

Added:
  stable/11/secure/lib/libcrypto/opensslconf-aarch64.h.in
 - copied unchanged from r304320, 
head/secure/lib/libcrypto/opensslconf-aarch64.h.in
  stable/11/secure/lib/libcrypto/opensslconf-arm.h.in
 - copied unchanged from r304320, 
head/secure/lib/libcrypto/opensslconf-arm.h.in
  stable/11/secure/lib/libcrypto/opensslconf-mips.h.in
 - copied unchanged from r304320, 
head/secure/lib/libcrypto/opensslconf-mips.h.in
  stable/11/secure/lib/libcrypto/opensslconf-powerpc.h.in
 - copied unchanged from r304320, 
head/secure/lib/libcrypto/opensslconf-powerpc.h.in
  stable/11/secure/lib/libcrypto/opensslconf-riscv.h.in
 - copied unchanged from r304320, 
head/secure/lib/libcrypto/opensslconf-riscv.h.in
  stable/11/secure/lib/libcrypto/opensslconf-sparc64.h.in
 - copied unchanged from r304320, 
head/secure/lib/libcrypto/opensslconf-sparc64.h.in
  stable/11/secure/lib/libcrypto/opensslconf-x86.h.in
 - copied unchanged from r304320, 
head/secure/lib/libcrypto/opensslconf-x86.h.in
Deleted:
  stable/11/secure/lib/libcrypto/opensslconf-aarch64.h
  stable/11/secure/lib/libcrypto/opensslconf-arm.h
  stable/11/secure/lib/libcrypto/opensslconf-mips.h
  stable/11/secure/lib/libcrypto/opensslconf-powerpc.h
  stable/11/secure/lib/libcrypto/opensslconf-riscv.h
  stable/11/secure/lib/libcrypto/opensslconf-sparc64.h
  stable/11/secure/lib/libcrypto/opensslconf-x86.h
Modified:
  stable/11/secure/lib/libcrypto/Makefile
  stable/11/secure/lib/libcrypto/Makefile.asm
  stable/11/secure/lib/libcrypto/Makefile.inc
Directory Properties:
  stable/11/   (props changed)

Modified: stable/11/secure/lib/libcrypto/Makefile
==
--- stable/11/secure/lib/libcrypto/Makefile Wed Aug 31 19:37:10 2016
(r305149)
+++ stable/11/secure/lib/libcrypto/Makefile Wed Aug 31 20:21:03 2016
(r305150)
@@ -22,9 +22,9 @@ MAN+= config.5 des_modes.7
 # base sources
 SRCS=  cpt_err.c cryptlib.c cversion.c ex_data.c mem.c mem_dbg.c o_dir.c \
o_fips.c o_init.c o_str.c o_time.c uid.c
-.if ${MACHINE_CPUARCH} == "amd64"
+.if defined(ASM_amd64)
 SRCS+= x86_64cpuid.S
-.elif ${MACHINE_CPUARCH} == "i386"
+.elif defined(ASM_i386)
 SRCS+= x86cpuid.S
 .else
 SRCS+= mem_clr.c
@@ -33,10 +33,10 @@ INCS+=  crypto.h ebcdic.h opensslv.h ossl
 
 # aes
 SRCS+= aes_cfb.c aes_ctr.c aes_ecb.c aes_ige.c aes_misc.c aes_ofb.c aes_wrap.c
-.if ${MACHINE_CPUARCH} == "amd64"
+.if defined(ASM_amd64)
 SRCS+= aes-x86_64.S aesni-mb-x86_64.S aesni-sha1-x86_64.S \
aesni-sha256-x86_64.S aesni-x86_64.S bsaes-x86_64.S vpaes-x86_64.S
-.elif ${MACHINE_CPUARCH} == "i386"
+.elif defined(ASM_i386)
 SRCS+= aes-586.S aesni-x86.S vpaes-x86.S
 .else
 SRCS+= aes_cbc.c aes_core.c
@@ -60,7 +60,7 @@ INCS+=asn1.h asn1_mac.h asn1t.h
 
 # bf
 SRCS+= bf_cfb64.c bf_ecb.c bf_ofb64.c bf_skey.c
-.if ${MACHINE_CPUARCH} == "i386"
+.if defined(ASM_i386)
 .if ${MACHINE_CPU:Mi686}
 SRCS+= bf-686.S
 .else
@@ -82,10 +82,10 @@ SRCS+=  bn_add.c bn_blind.c bn_const.c bn
bn_exp.c bn_exp2.c bn_gcd.c bn_gf2m.c bn_kron.c bn_lib.c bn_mod.c \
bn_mont.c bn_mpi.c bn_mul.c bn_nist.c bn_prime.c bn_print.c bn_rand.c \
bn_recp.c bn_shift.c bn_sqr.c bn_sqrt.c bn_word.c bn_x931p.c
-.if ${MACHINE_CPUARCH} == "amd64"
+.if defined(ASM_amd64)
 SRCS+= rsaz-avx2.S rsaz-x86_64.S rsaz_exp.c x86_64-gcc.c x86_64-gf2m.S \
x86_64-mont.S x86_64-mont5.S
-.elif ${MACHINE_CPUARCH} == "i386"
+.elif defined(ASM_i386)
 SRCS+= bn-586.S co-586.S x86-gf2m.S x86-mont.S
 .else
 SRCS+= bn_asm.c
@@ -98,9 +98,9 @@ INCS+=buffer.h
 
 # camellia
 SRCS+= cmll_cfb.c cmll_ctr.c cmll_ecb.c cmll_ofb.c cmll_utl.c
-.if ${MACHINE_CPUARCH} == "amd64"
+.if defined(ASM_amd64)
 SRCS+= cmll_misc.c cmll-x86_64.S
-.elif ${MACHINE_CPUARCH} == "i386"
+.elif defined(ASM_i386)
 SRCS+= cmll-x86.S
 .else
 SRCS+= camellia.c cmll_cbc.c cmll_misc.c
@@ -135,7 +135,7 @@ SRCS+=  cbc_cksm.c cbc_enc.c cfb64ede.c c
des_old2.c ecb3_enc.c ecb_enc.c ede_cbcm_enc.c enc_read.c enc_writ.c \
fcrypt.c ofb64ede.c ofb64enc.c ofb_enc.c pcbc_enc.c qud_cksm.c \
rand_key.c read2pwd.c rpc_enc.c set_key.c str2key.c xcbc_enc.c
-.if ${MACHINE_CPUARCH} == "i386"
+.if defined(ASM_i386)
 SRCS+= crypt586.S des-586.S
 .else
 SRCS+= des_enc.c fcrypt_b.c
@@ -161,7 +161,7 @@ SRCS+=  ec2_mult.c ec2_oct.c ec2_smpl.c e
ec_curve.c ec_cvt.c ec_err.c ec_key.c ec_lib.c ec_mult.c ec_oct.c \
ec_pmeth.c ec_print.c eck_prn.c ecp_mont.c ecp_nist.c ecp_oct.c \
ecp_smpl.c
-.if ${MACHINE_CPUARCH} == "amd64"
+.if defined(ASM_amd64)
 SRCS+= ecp_nistz256.c ecp_nistz256-x86_64.S
 .endif

Re: svn commit: r305114 - head/sys/arm/ti

2016-08-31 Thread Luiz Otavio O Souza 
On Wed, Aug 31, 2016 at 3:54 PM, wrote:
>
>> On Aug 31, 2016, at 12:22 AM, Luiz Otavio O Souza wrote:
>>
>> Author: loos
>> Date: Wed Aug 31 07:22:14 2016
>> New Revision: 305114
>> URL: https://svnweb.freebsd.org/changeset/base/305114
>>
>> Log:
>>  Allow the use of control module extensions to cope with specific platform
>>  features.
>>
>>  Sponsored by:Rubicon Communications, LLC (Netgate)
>>
>> Modified:
>>  head/sys/arm/ti/ti_scm.c
>>
>> Modified: head/sys/arm/ti/ti_scm.c
>> ==
>> --- head/sys/arm/ti/ti_scm.c  Wed Aug 31 07:03:06 2016(r305113)
>> +++ head/sys/arm/ti/ti_scm.c  Wed Aug 31 07:22:14 2016(r305114)
>> @@ -131,7 +131,11 @@ ti_scm_attach(device_t dev)
>>
>>   ti_scm_sc = sc;
>>
>> - return (0);
>> + /* Attach platform extensions, if any. */
>> + bus_generic_probe(dev);
>> + bus_enumerate_hinted_children(dev);
>> +
>> + return (bus_generic_attach(dev));
>> }
>
> Is bus_enumerate_hinted_children really required? I think all TI platforms 
> are FDT-compatible


No, it is not required.  Fixed in r305149.

Thanks!

Luiz
___
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: r305149 - head/sys/arm/ti

2016-08-31 Thread Luiz Otavio O Souza 
Author: loos
Date: Wed Aug 31 19:37:10 2016
New Revision: 305149
URL: https://svnweb.freebsd.org/changeset/base/305149

Log:
  All the TI platforms are FDT based and it is not necessary to enumerate
  the hinted children (there wont be any).
  
  Spotted by:   gonzo

Modified:
  head/sys/arm/ti/ti_scm.c

Modified: head/sys/arm/ti/ti_scm.c
==
--- head/sys/arm/ti/ti_scm.cWed Aug 31 19:30:59 2016(r305148)
+++ head/sys/arm/ti/ti_scm.cWed Aug 31 19:37:10 2016(r305149)
@@ -133,7 +133,6 @@ ti_scm_attach(device_t dev)
 
/* Attach platform extensions, if any. */
bus_generic_probe(dev);
-   bus_enumerate_hinted_children(dev);
 
return (bus_generic_attach(dev));
 }
___
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: r305147 - in head: . share/mk

2016-08-31 Thread Bryan Drewery 
Author: bdrewery
Date: Wed Aug 31 19:30:52 2016
New Revision: 305147
URL: https://svnweb.freebsd.org/changeset/base/305147

Log:
  Add a 'make print-dir' that simply traverses all directories and prints them.
  
  This is useful for finding connected directories.
  
  MFC after:1 week
  Sponsored by: EMC / Isilon Storage Division

Modified:
  head/Makefile
  head/share/mk/bsd.dep.mk
  head/share/mk/bsd.init.mk
  head/share/mk/bsd.subdir.mk
  head/share/mk/local.meta.sys.mk
  head/share/mk/sys.mk

Modified: head/Makefile
==
--- head/Makefile   Wed Aug 31 19:30:46 2016(r305146)
+++ head/Makefile   Wed Aug 31 19:30:52 2016(r305147)
@@ -109,7 +109,8 @@
 
 # Note: we use this awkward construct to be compatible with FreeBSD's
 # old make used in 10.0 and 9.2 and earlier.
-.if defined(MK_DIRDEPS_BUILD) && ${MK_DIRDEPS_BUILD} == "yes" && 
!make(showconfig)
+.if defined(MK_DIRDEPS_BUILD) && ${MK_DIRDEPS_BUILD} == "yes" && \
+!make(showconfig) && !make(print-dir)
 # targets/Makefile plays the role of top-level
 .include "targets/Makefile"
 .else
@@ -132,7 +133,7 @@ TGTS=   all all-man buildenv buildenvvars 
xdev-links native-xtools stageworld stagekernel stage-packages \
create-world-packages create-kernel-packages create-packages \
packages installconfig real-packages sign-packages package-pkg \
-   test-system-compiler
+   print-dir test-system-compiler
 
 # XXX: r156740: This can't work since bsd.subdir.mk is not included ever.
 # It will only work for SUBDIR_TARGETS in make.conf.
@@ -257,6 +258,10 @@ _TARGET_ARCH=  ${XDEV_ARCH}
 _TARGET?=  ${MACHINE}
 _TARGET_ARCH?= ${MACHINE_ARCH}
 
+.if make(print-dir)
+.SILENT:
+.endif
+
 #
 # Make sure we have an up-to-date make(1). Only world and buildworld
 # should do this as those are the initial targets used for upgrades.

Modified: head/share/mk/bsd.dep.mk
==
--- head/share/mk/bsd.dep.mkWed Aug 31 19:30:46 2016(r305146)
+++ head/share/mk/bsd.dep.mkWed Aug 31 19:30:52 2016(r305147)
@@ -82,9 +82,9 @@ _meta_filemon=1
 # since it will track dependencies itself.  OBJS_DEPEND_GUESS is still used.
 .if !empty(.MAKEFLAGS:M-V${_V_READ_DEPEND}) || make(obj) || make(clean*) || \
 ${.TARGETS:M*install*} == ${.TARGETS} || \
-make(analyze) || defined(_meta_filemon)
+make(analyze) || defined(_meta_filemon) || make(print-dir)
 _SKIP_READ_DEPEND= 1
-.if ${MK_DIRDEPS_BUILD} == "no"
+.if ${MK_DIRDEPS_BUILD} == "no" || make(analyze) || make(print-dir)
 .MAKE.DEPENDFILE=  /dev/null
 .endif
 .endif
@@ -198,7 +198,7 @@ CFLAGS+=${DEPEND_CFLAGS}
 .endif # !defined(_meta_filemon)
 .endif # defined(SRCS)
 
-.if ${MK_DIRDEPS_BUILD} == "yes"
+.if ${MK_DIRDEPS_BUILD} == "yes" && !make(analyze) && !make(print-dir)
 # Prevent meta.autodep.mk from tracking "local dependencies".
 .depend:
 .include 

Modified: head/share/mk/bsd.init.mk
==
--- head/share/mk/bsd.init.mk   Wed Aug 31 19:30:46 2016(r305146)
+++ head/share/mk/bsd.init.mk   Wed Aug 31 19:30:52 2016(r305147)
@@ -31,7 +31,7 @@ _SKIP_BUILD = not building at level 0
 .if ${MK_META_MODE} == "yes"
 .if !exists(/dev/filemon) && \
 ${UPDATE_DEPENDFILE:Uyes:tl} != "no" && !defined(NO_FILEMON) && \
-!make(showconfig) && ${.MAKEFLAGS:M-V} == ""
+!make(showconfig) && !make(print-dir) && ${.MAKEFLAGS:M-V} == ""
 .warning The filemon module (/dev/filemon) is not loaded.
 .warning META_MODE is less useful for incremental builds without filemon.
 .warning 'kldload filemon' or pass -DNO_FILEMON to suppress this warning.

Modified: head/share/mk/bsd.subdir.mk
==
--- head/share/mk/bsd.subdir.mk Wed Aug 31 19:30:46 2016(r305146)
+++ head/share/mk/bsd.subdir.mk Wed Aug 31 19:30:52 2016(r305147)
@@ -42,15 +42,15 @@ SUBDIR_TARGETS+= \
all all-man analyze buildconfig buildfiles buildincludes \
checkdpadd clean cleandepend cleandir cleanilinks \
cleanobj depend distribute files includes installconfig \
-   installfiles installincludes realinstall lint maninstall \
-   manlint obj objlink tags \
+   installfiles installincludes print-dir realinstall lint \
+   maninstall manlint obj objlink tags \
 
 # Described above.
 STANDALONE_SUBDIR_TARGETS+= \
all-man buildconfig buildfiles buildincludes check checkdpadd \
clean cleandepend cleandir cleanilinks cleanobj files includes \
-   installconfig installincludes installfiles maninstall manlint \
-   obj objlink \
+   installconfig installincludes installfiles print-dir \
+

svn commit: r305146 - in head: cddl/usr.sbin/dtrace cddl/usr.sbin/lockstat cddl/usr.sbin/plockstat lib/clang/libclangformat lib/clang/libclangtoolingcore lib/libc/tests/stdlib lib/libproc lib/libpr...

2016-08-31 Thread Bryan Drewery 
Author: bdrewery
Date: Wed Aug 31 19:30:46 2016
New Revision: 305146
URL: https://svnweb.freebsd.org/changeset/base/305146

Log:
  DIRDEPS_BUILD: Update dependencies.
  
  Sponsored by: EMC / Isilon Storage Division

Modified:
  head/cddl/usr.sbin/dtrace/Makefile.depend
  head/cddl/usr.sbin/lockstat/Makefile.depend
  head/cddl/usr.sbin/plockstat/Makefile.depend
  head/lib/clang/libclangformat/Makefile.depend
  head/lib/clang/libclangtoolingcore/Makefile.depend
  head/lib/libc/tests/stdlib/Makefile.depend
  head/lib/libproc/Makefile.depend
  head/lib/libproc/tests/Makefile.depend
  head/lib/librtld_db/Makefile.depend
  head/secure/usr.bin/ssh/Makefile.depend
  head/usr.bin/nfsstat/Makefile.depend
  head/usr.bin/tar/tests/Makefile.depend
  head/usr.sbin/bhyve/Makefile.depend
  head/usr.sbin/pw/tests/Makefile.depend

Modified: head/cddl/usr.sbin/dtrace/Makefile.depend
==
--- head/cddl/usr.sbin/dtrace/Makefile.depend   Wed Aug 31 19:30:00 2016
(r305145)
+++ head/cddl/usr.sbin/dtrace/Makefile.depend   Wed Aug 31 19:30:46 2016
(r305146)
@@ -13,7 +13,9 @@ DIRDEPS = \
lib/libcompiler_rt \
lib/libcxxrt \
lib/libelf \
+   lib/libkvm \
lib/libproc \
+   lib/libprocstat \
lib/librtld_db \
lib/libthr \
lib/libutil \

Modified: head/cddl/usr.sbin/lockstat/Makefile.depend
==
--- head/cddl/usr.sbin/lockstat/Makefile.depend Wed Aug 31 19:30:00 2016
(r305145)
+++ head/cddl/usr.sbin/lockstat/Makefile.depend Wed Aug 31 19:30:46 2016
(r305146)
@@ -13,7 +13,9 @@ DIRDEPS = \
lib/libcompiler_rt \
lib/libcxxrt \
lib/libelf \
+   lib/libkvm \
lib/libproc \
+   lib/libprocstat \
lib/librt \
lib/librtld_db \
lib/libthr \

Modified: head/cddl/usr.sbin/plockstat/Makefile.depend
==
--- head/cddl/usr.sbin/plockstat/Makefile.dependWed Aug 31 19:30:00 
2016(r305145)
+++ head/cddl/usr.sbin/plockstat/Makefile.dependWed Aug 31 19:30:46 
2016(r305146)
@@ -13,7 +13,9 @@ DIRDEPS = \
lib/libcompiler_rt \
lib/libcxxrt \
lib/libelf \
+   lib/libkvm \
lib/libproc \
+   lib/libprocstat \
lib/librtld_db \
lib/libthr \
lib/libutil \

Modified: head/lib/clang/libclangformat/Makefile.depend
==
--- head/lib/clang/libclangformat/Makefile.depend   Wed Aug 31 19:30:00 
2016(r305145)
+++ head/lib/clang/libclangformat/Makefile.depend   Wed Aug 31 19:30:46 
2016(r305146)
@@ -6,6 +6,7 @@ DIRDEPS = \
include/xlocale \
lib/libc++ \
lib/msun \
+   usr.bin/clang/clang-tblgen.host \
 
 
 .include 

Modified: head/lib/clang/libclangtoolingcore/Makefile.depend
==
--- head/lib/clang/libclangtoolingcore/Makefile.depend  Wed Aug 31 19:30:00 
2016(r305145)
+++ head/lib/clang/libclangtoolingcore/Makefile.depend  Wed Aug 31 19:30:46 
2016(r305146)
@@ -6,6 +6,7 @@ DIRDEPS = \
include/xlocale \
lib/libc++ \
lib/msun \
+   usr.bin/clang/clang-tblgen.host \
 
 
 .include 

Modified: head/lib/libc/tests/stdlib/Makefile.depend
==
--- head/lib/libc/tests/stdlib/Makefile.depend  Wed Aug 31 19:30:00 2016
(r305145)
+++ head/lib/libc/tests/stdlib/Makefile.depend  Wed Aug 31 19:30:46 2016
(r305146)
@@ -11,9 +11,10 @@ DIRDEPS = \
lib/atf/libatf-c++ \
lib/libc \
lib/libc++ \
-   lib/libthr \
lib/libcompiler_rt \
+   lib/libcxxrt \
lib/libnetbsd \
+   lib/libthr \
lib/libutil \
lib/msun \
 

Modified: head/lib/libproc/Makefile.depend
==
--- head/lib/libproc/Makefile.dependWed Aug 31 19:30:00 2016
(r305145)
+++ head/lib/libproc/Makefile.dependWed Aug 31 19:30:46 2016
(r305146)
@@ -12,6 +12,7 @@ DIRDEPS = \
lib/libcompiler_rt \
lib/libcxxrt \
lib/libelf \
+   lib/libprocstat \
lib/librtld_db \
lib/libutil \
 

Modified: head/lib/libproc/tests/Makefile.depend
==
--- head/lib/libproc/tests/Makefile.depend  Wed Aug 31 19:30:00 2016
(r305145)
+++ head/lib/libproc/tests/Makefile.depend  Wed Aug 31 19:30:46 2016
(r305146)
@@ -13,7 +13,9 @@ DIRDEPS = \
lib/libcompiler_rt \
lib/libcxxrt \
lib/libelf \
+   lib/libkvm \
lib/libproc \
+

svn commit: r305148 - in head: cddl/usr.sbin/zfsd libexec/rtld-elf/tests/libpythagoras libexec/rtld-elf/tests/target targets/pseudo/tests targets/pseudo/userland targets/pseudo/userland/cddl target...

2016-08-31 Thread Bryan Drewery 
Author: bdrewery
Date: Wed Aug 31 19:30:59 2016
New Revision: 305148
URL: https://svnweb.freebsd.org/changeset/base/305148

Log:
  DIRDEPS_BUILD: Add some missing dirctories to the build.
  
  Sponsored by: EMC / Isilon Storage Division

Added:
  head/cddl/usr.sbin/zfsd/Makefile.depend   (contents, props changed)
  head/libexec/rtld-elf/tests/libpythagoras/Makefile.depend   (contents, props 
changed)
  head/libexec/rtld-elf/tests/target/Makefile.depend   (contents, props changed)
  head/tests/sys/pjdfstest/pjdfstest/Makefile.depend   (contents, props changed)
  head/tests/sys/sys/Makefile.depend   (contents, props changed)
  head/usr.bin/bsdcat/tests/Makefile.depend   (contents, props changed)
  head/usr.bin/sdiff/tests/Makefile.depend   (contents, props changed)
  head/usr.bin/xinstall/tests/Makefile.depend   (contents, props changed)
  head/usr.sbin/extattr/tests/Makefile.depend   (contents, props changed)
  head/usr.sbin/pmcstudy/Makefile.depend   (contents, props changed)
  head/usr.sbin/ypldap/Makefile.depend   (contents, props changed)
Modified:
  head/targets/pseudo/tests/Makefile.depend
  head/targets/pseudo/userland/Makefile.depend
  head/targets/pseudo/userland/cddl/Makefile.depend
  head/targets/pseudo/userland/libexec/Makefile.depend

Added: head/cddl/usr.sbin/zfsd/Makefile.depend
==
--- /dev/null   00:00:00 1970   (empty, because file is newly added)
+++ head/cddl/usr.sbin/zfsd/Makefile.depend Wed Aug 31 19:30:59 2016
(r305148)
@@ -0,0 +1,35 @@
+# $FreeBSD$
+# Autogenerated - do NOT edit!
+
+DIRDEPS = \
+   cddl/lib/libavl \
+   cddl/lib/libnvpair \
+   cddl/lib/libumem \
+   cddl/lib/libuutil \
+   cddl/lib/libzfs \
+   cddl/lib/libzfs_core \
+   gnu/lib/csu \
+   gnu/lib/libgcc \
+   include \
+   include/xlocale \
+   lib/${CSU_DIR} \
+   lib/libc \
+   lib/libc++ \
+   lib/libcompiler_rt \
+   lib/libcxxrt \
+   lib/libdevdctl \
+   lib/libexpat \
+   lib/libgeom \
+   lib/libmd \
+   lib/libsbuf \
+   lib/libthr \
+   lib/libutil \
+   lib/libz \
+   lib/msun \
+
+
+.include 
+
+.if ${DEP_RELDIR} == ${_DEP_RELDIR}
+# local dependencies - needed for -jN in clean tree
+.endif

Added: head/libexec/rtld-elf/tests/libpythagoras/Makefile.depend
==
--- /dev/null   00:00:00 1970   (empty, because file is newly added)
+++ head/libexec/rtld-elf/tests/libpythagoras/Makefile.depend   Wed Aug 31 
19:30:59 2016(r305148)
@@ -0,0 +1,18 @@
+# $FreeBSD$
+# Autogenerated - do NOT edit!
+
+DIRDEPS = \
+   gnu/lib/csu \
+   gnu/lib/libgcc \
+   include \
+   lib/${CSU_DIR} \
+   lib/libc \
+   lib/libcompiler_rt \
+   lib/msun \
+
+
+.include 
+
+.if ${DEP_RELDIR} == ${_DEP_RELDIR}
+# local dependencies - needed for -jN in clean tree
+.endif

Added: head/libexec/rtld-elf/tests/target/Makefile.depend
==
--- /dev/null   00:00:00 1970   (empty, because file is newly added)
+++ head/libexec/rtld-elf/tests/target/Makefile.depend  Wed Aug 31 19:30:59 
2016(r305148)
@@ -0,0 +1,18 @@
+# $FreeBSD$
+# Autogenerated - do NOT edit!
+
+DIRDEPS = \
+   gnu/lib/csu \
+   gnu/lib/libgcc \
+   include \
+   lib/${CSU_DIR} \
+   lib/libc \
+   lib/libcompiler_rt \
+   libexec/rtld-elf/tests/libpythagoras \
+
+
+.include 
+
+.if ${DEP_RELDIR} == ${_DEP_RELDIR}
+# local dependencies - needed for -jN in clean tree
+.endif

Modified: head/targets/pseudo/tests/Makefile.depend
==
--- head/targets/pseudo/tests/Makefile.depend   Wed Aug 31 19:30:52 2016
(r305147)
+++ head/targets/pseudo/tests/Makefile.depend   Wed Aug 31 19:30:59 2016
(r305148)
@@ -166,6 +166,8 @@ DIRDEPS= \
libexec/atf/atf-sh/tests \
libexec/atf/tests \
libexec/rtld-elf/tests \
+   libexec/rtld-elf/tests/libpythagoras \
+   libexec/rtld-elf/tests/target \
libexec/tests \
sbin/devd/tests \
sbin/dhclient/tests \
@@ -213,6 +215,7 @@ DIRDEPS= \
tests/sys/mqueue \
tests/sys/netinet \
tests/sys/opencrypto \
+   tests/sys/pjdfstest/pjdfstest \
tests/sys/pjdfstest/tests \
tests/sys/pjdfstest/tests/chflags \
tests/sys/pjdfstest/tests/chmod \
@@ -230,6 +233,7 @@ DIRDEPS= \
tests/sys/pjdfstest/tests/truncate \
tests/sys/pjdfstest/tests/unlink \
tests/sys/posixshm \
+   tests/sys/sys \
tests/sys/vfs \
tests/sys/vm \
usr.bin/apply/tests \
@@ -303,6 +307,7 @@ DIRDEPS= \
usr.bin/mkimg/tests \
usr.bin/ncal/tests \
usr.bin/printf/tests \
+   usr.bin/sdiff/tests \
usr.bin/sed/tests \

svn commit: r305145 - head/gnu/lib/libgcc

2016-08-31 Thread Bryan Drewery 
Author: bdrewery
Date: Wed Aug 31 19:30:00 2016
New Revision: 305145
URL: https://svnweb.freebsd.org/changeset/base/305145

Log:
  DIRDEPS_BUILD: Avoid cyclic dependency with libc++.
  
  The DIRDEPS_BUILD does not have a 'make includes' phase, so it would
  otherwise want libc++ to be fully built/staged before building
  libgcc.  Using the header directly works.
  
  MFC after:3 days
  Sponsored by: EMC / Isilon Storage Division

Modified:
  head/gnu/lib/libgcc/Makefile

Modified: head/gnu/lib/libgcc/Makefile
==
--- head/gnu/lib/libgcc/MakefileWed Aug 31 18:49:50 2016
(r305144)
+++ head/gnu/lib/libgcc/MakefileWed Aug 31 19:30:00 2016
(r305145)
@@ -95,6 +95,10 @@ CXXFLAGS+=   -std=c++11
 .endif
 CXXFLAGS+= -fno-rtti
 STATIC_CXXFLAGS+= -fvisibility=hidden -fPIC
+.if ${MK_DIRDEPS_BUILD} == "yes"
+# Avoid dependency on lib/libc++
+CFLAGS+=   -I${SRCTOP}/contrib/libc++/include
+.endif
 
 .else # MK_LLVM_LIBUNWIND
 
___
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: r305114 - head/sys/arm/ti

2016-08-31 Thread Oleksandr Tymoshenko 

> On Aug 31, 2016, at 12:22 AM, Luiz Otavio O Souza  wrote:
> 
> Author: loos
> Date: Wed Aug 31 07:22:14 2016
> New Revision: 305114
> URL: https://svnweb.freebsd.org/changeset/base/305114
> 
> Log:
>  Allow the use of control module extensions to cope with specific platform
>  features.
> 
>  Sponsored by:Rubicon Communications, LLC (Netgate)
> 
> Modified:
>  head/sys/arm/ti/ti_scm.c
> 
> Modified: head/sys/arm/ti/ti_scm.c
> ==
> --- head/sys/arm/ti/ti_scm.c  Wed Aug 31 07:03:06 2016(r305113)
> +++ head/sys/arm/ti/ti_scm.c  Wed Aug 31 07:22:14 2016(r305114)
> @@ -131,7 +131,11 @@ ti_scm_attach(device_t dev)
> 
>   ti_scm_sc = sc;
> 
> - return (0);
> + /* Attach platform extensions, if any. */
> + bus_generic_probe(dev);
> + bus_enumerate_hinted_children(dev);
> +
> + return (bus_generic_attach(dev));
> }

Is bus_enumerate_hinted_children really required? I think all TI platforms are 
FDT-compatible 
___
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: r305144 - head/lib/libc/net

2016-08-31 Thread Andrey A. Chernov 
Author: ache
Date: Wed Aug 31 18:49:50 2016
New Revision: 305144
URL: https://svnweb.freebsd.org/changeset/base/305144

Log:
  'addrlen' does not matter when we need to find the first non-zero bit in
  the byte from the left and 'addrlen' already counted in 'lim'.
  
  PR: 212121
  Submitted by:   herbie.robin...@stratus.com
  MFC after:  7 days

Modified:
  head/lib/libc/net/getaddrinfo.c
  head/lib/libc/net/name6.c

Modified: head/lib/libc/net/getaddrinfo.c
==
--- head/lib/libc/net/getaddrinfo.c Wed Aug 31 18:37:51 2016
(r305143)
+++ head/lib/libc/net/getaddrinfo.c Wed Aug 31 18:49:50 2016
(r305144)
@@ -949,7 +949,7 @@ matchlen(struct sockaddr *src, struct so
 
while (s < lim)
if ((r = (*d++ ^ *s++)) != 0) {
-   while (r < addrlen * 8) {
+   while ((r & 0x80) == 0) {
match++;
r <<= 1;
}

Modified: head/lib/libc/net/name6.c
==
--- head/lib/libc/net/name6.c   Wed Aug 31 18:37:51 2016(r305143)
+++ head/lib/libc/net/name6.c   Wed Aug 31 18:49:50 2016(r305144)
@@ -930,7 +930,7 @@ matchlen(struct sockaddr *src, struct so
 
while (s < lim)
if ((r = (*d++ ^ *s++)) != 0) {
-   while (r < addrlen * 8) {
+   while ((r & 0x80) == 0) {
match++;
r <<= 1;
}
___
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: r305143 - head/sys/kern

2016-08-31 Thread Ed Maste 
Author: emaste
Date: Wed Aug 31 18:37:51 2016
New Revision: 305143
URL: https://svnweb.freebsd.org/changeset/base/305143

Log:
  Regnerate after r305140, getdtablesize in capability mode
  
  Sponsored by: The FreeBSD Foundation

Modified:
  head/sys/kern/init_sysent.c

Modified: head/sys/kern/init_sysent.c
==
--- head/sys/kern/init_sysent.c Wed Aug 31 18:37:04 2016(r305142)
+++ head/sys/kern/init_sysent.c Wed Aug 31 18:37:51 2016(r305143)
@@ -135,7 +135,7 @@ struct sysent sysent[] = {
{ AS(getitimer_args), (sy_call_t *)sys_getitimer, AUE_GETITIMER, NULL, 
0, 0, SYF_CAPENABLED, SY_THR_STATIC },   /* 86 = getitimer */
{ compat(AS(gethostname_args),gethostname), AUE_SYSCTL, NULL, 0, 0, 
SYF_CAPENABLED, SY_THR_STATIC },/* 87 = old gethostname */
{ compat(AS(sethostname_args),sethostname), AUE_SYSCTL, NULL, 0, 0, 0, 
SY_THR_STATIC }, /* 88 = old sethostname */
-   { 0, (sy_call_t *)sys_getdtablesize, AUE_GETDTABLESIZE, NULL, 0, 0, 0, 
SY_THR_STATIC }, /* 89 = getdtablesize */
+   { 0, (sy_call_t *)sys_getdtablesize, AUE_GETDTABLESIZE, NULL, 0, 0, 
SYF_CAPENABLED, SY_THR_STATIC },/* 89 = getdtablesize */
{ AS(dup2_args), (sy_call_t *)sys_dup2, AUE_DUP2, NULL, 0, 0, 
SYF_CAPENABLED, SY_THR_STATIC },  /* 90 = dup2 */
{ 0, (sy_call_t *)nosys, AUE_NULL, NULL, 0, 0, 0, SY_THR_ABSENT },  
/* 91 = getdopt */
{ AS(fcntl_args), (sy_call_t *)sys_fcntl, AUE_FCNTL, NULL, 0, 0, 
SYF_CAPENABLED, SY_THR_STATIC },   /* 92 = fcntl */
___
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: r305142 - in stable: 10/sbin/hastd 11/sbin/hastd

2016-08-31 Thread Dimitry Andric 
Author: dim
Date: Wed Aug 31 18:37:04 2016
New Revision: 305142
URL: https://svnweb.freebsd.org/changeset/base/305142

Log:
  MFC r304969:
  
  Define hastd's STRICT_ALIGN macro in a defined and portable way.

Modified:
  stable/10/sbin/hastd/lzf.h
Directory Properties:
  stable/10/   (props changed)

Changes in other areas also in this revision:
Modified:
  stable/11/sbin/hastd/lzf.h
Directory Properties:
  stable/11/   (props changed)

Modified: stable/10/sbin/hastd/lzf.h
==
--- stable/10/sbin/hastd/lzf.h  Wed Aug 31 18:35:41 2016(r305141)
+++ stable/10/sbin/hastd/lzf.h  Wed Aug 31 18:37:04 2016(r305142)
@@ -132,7 +132,11 @@ lzf_decompress (const void *const in_dat
  * Unconditionally aligning does not cost very much, so do it if unsure
  */
 #ifndef STRICT_ALIGN
-# define STRICT_ALIGN !(defined(__i386) || defined (__amd64))
+# if !(defined(__i386) || defined (__amd64))
+#  define STRICT_ALIGN 1
+# else
+#  define STRICT_ALIGN 0
+# endif
 #endif
 
 /*
___
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: r305142 - in stable: 10/sbin/hastd 11/sbin/hastd

2016-08-31 Thread Dimitry Andric 
Author: dim
Date: Wed Aug 31 18:37:04 2016
New Revision: 305142
URL: https://svnweb.freebsd.org/changeset/base/305142

Log:
  MFC r304969:
  
  Define hastd's STRICT_ALIGN macro in a defined and portable way.

Modified:
  stable/11/sbin/hastd/lzf.h
Directory Properties:
  stable/11/   (props changed)

Changes in other areas also in this revision:
Modified:
  stable/10/sbin/hastd/lzf.h
Directory Properties:
  stable/10/   (props changed)

Modified: stable/11/sbin/hastd/lzf.h
==
--- stable/11/sbin/hastd/lzf.h  Wed Aug 31 18:35:41 2016(r305141)
+++ stable/11/sbin/hastd/lzf.h  Wed Aug 31 18:37:04 2016(r305142)
@@ -132,7 +132,11 @@ lzf_decompress (const void *const in_dat
  * Unconditionally aligning does not cost very much, so do it if unsure
  */
 #ifndef STRICT_ALIGN
-# define STRICT_ALIGN !(defined(__i386) || defined (__amd64))
+# if !(defined(__i386) || defined (__amd64))
+#  define STRICT_ALIGN 1
+# else
+#  define STRICT_ALIGN 0
+# endif
 #endif
 
 /*
___
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: r305141 - head/sys/arm/ti/am335x

2016-08-31 Thread Luiz Otavio O Souza 
Author: loos
Date: Wed Aug 31 18:35:41 2016
New Revision: 305141
URL: https://svnweb.freebsd.org/changeset/base/305141

Log:
  Fix the build.
  
  Works better when the file with the #define is actually included...

Modified:
  head/sys/arm/ti/am335x/am335x_prcm.c

Modified: head/sys/arm/ti/am335x/am335x_prcm.c
==
--- head/sys/arm/ti/am335x/am335x_prcm.cWed Aug 31 18:33:15 2016
(r305140)
+++ head/sys/arm/ti/am335x/am335x_prcm.cWed Aug 31 18:35:41 2016
(r305141)
@@ -52,6 +52,8 @@ __FBSDID("$FreeBSD$");
 
 #include 
 
+#include "am335x_scm.h"
+
 #define CM_PER 0
 #define CM_PER_L4LS_CLKSTCTRL  (CM_PER + 0x000)
 #define CM_PER_L3S_CLKSTCTRL   (CM_PER + 0x004)
___
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: r305140 - head/sys/kern

2016-08-31 Thread Ed Maste 
Author: emaste
Date: Wed Aug 31 18:33:15 2016
New Revision: 305140
URL: https://svnweb.freebsd.org/changeset/base/305140

Log:
  Allow getdtablesize in capability mode
  
  getdtablesize is "trivial global state" and is similar to
  getrlimit(RLIMIT_NOFILE), so should be permitted in capability mode.
  
  Reviewed by:  oshogbo
  MFC after:1 week
  Sponsored by: The FreeBSD Foundation
  Differential Revision:https://reviews.freebsd.org/D7719

Modified:
  head/sys/kern/capabilities.conf

Modified: head/sys/kern/capabilities.conf
==
--- head/sys/kern/capabilities.conf Wed Aug 31 18:10:41 2016
(r305139)
+++ head/sys/kern/capabilities.conf Wed Aug 31 18:33:15 2016
(r305140)
@@ -248,6 +248,7 @@ getdirentries
 ## Allow querying certain trivial global state.
 ##
 getdomainname
+getdtablesize
 
 ##
 ## Allow querying current process credential state.
___
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: r305107 - head/sys/boot/forth

2016-08-31 Thread Devin Teske 

> On Aug 31, 2016, at 7:43 AM, Warner Losh  wrote:
> 
> On Wed, Aug 31, 2016 at 8:41 AM, Devin Teske  wrote:
>> Please remove the following:
>> 
>> "Unsure why we don't just have start"
>> 
>> Others beside yourself have great understanding of the Forth and we don't 
>> need questioning comments in the code.
> 
> To be fair, it was just left in there accidentally. However, could you
> explain it to me? That answer is insulting without being actually
> helpful in resolving the situation.
> 

Many apologies, I did not mean to be insulting -- just under the gun and short 
on time for eloquence (poor excuse, I know).

Let me try and put something together that better answers any question(s) you 
may have about the Forth.

If you have some time later, maybe we can catch up on IRC and we can either 
improve the code, the comments, or understanding thereof not only for ourselves 
but for others as well.
-- 
Cheers,
Devin
___
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: r305139 - head/bin/df

2016-08-31 Thread Conrad E. Meyer 
Author: cem
Date: Wed Aug 31 18:10:41 2016
New Revision: 305139
URL: https://svnweb.freebsd.org/changeset/base/305139

Log:
  df(1): Allow duplicate -l flags gracefully
  
  Rather than producing a misleading error message when duplicate -l flags are
  provided to df(1), simply ignore extra flags and proceed as if only one was
  specified.  This seems most reasonable given the usage for -l:
  
   -l  Only display information about locally-mounted file systems.
  
  l and t flags still conflict, as before.
  
  PR:   208169
  Reported by:  by at reorigin.com
  Reviewed by:  allanjude

Modified:
  head/bin/df/df.c

Modified: head/bin/df/df.c
==
--- head/bin/df/df.cWed Aug 31 18:00:41 2016(r305138)
+++ head/bin/df/df.cWed Aug 31 18:10:41 2016(r305139)
@@ -166,6 +166,9 @@ main(int argc, char *argv[])
hflag = 0;
break;
case 'l':
+   /* Ignore duplicate -l */
+   if (lflag)
+   break;
if (vfslist != NULL)
xo_errx(1, "-l and -t are mutually exclusive.");
vfslist = makevfslist(makenetvfslist());
___
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: r305138 - in stable: 10/contrib/ipfilter 10/sys/contrib/ipfilter/netinet 11/contrib/ipfilter 11/sys/contrib/ipfilter/netinet

2016-08-31 Thread Dimitry Andric 
Author: dim
Date: Wed Aug 31 18:00:41 2016
New Revision: 305138
URL: https://svnweb.freebsd.org/changeset/base/305138

Log:
  MFC r304953:
  
  Define ipfilter's SOLARIS macro in a defined and portable way.
  
  Reviewed by:  cy
  Differential Revision: https://reviews.freebsd.org/D7671
  
  MFC r304959 (by kib):
  
  Complete r304953.
  
  Sponsored by: The FreeBSD Foundation
  
  MFC r304964:
  
  Follow-up to r304953, in which I broke the build: apparently the SOLARIS
  macro is defined in lots of different places in ipfilter, so replace all
  of the nonportable definitions with portable ones.
  
  Pointy hat to:dim

Modified:
  stable/10/contrib/ipfilter/opts.h
  stable/10/sys/contrib/ipfilter/netinet/ip_compat.h
  stable/10/sys/contrib/ipfilter/netinet/ip_fil.h
  stable/10/sys/contrib/ipfilter/netinet/ip_log.c
  stable/10/sys/contrib/ipfilter/netinet/ip_nat.h
  stable/10/sys/contrib/ipfilter/netinet/ip_proxy.h
Directory Properties:
  stable/10/   (props changed)

Changes in other areas also in this revision:
Modified:
  stable/11/contrib/ipfilter/opts.h
  stable/11/sys/contrib/ipfilter/netinet/ip_compat.h
  stable/11/sys/contrib/ipfilter/netinet/ip_fil.h
  stable/11/sys/contrib/ipfilter/netinet/ip_log.c
  stable/11/sys/contrib/ipfilter/netinet/ip_nat.h
  stable/11/sys/contrib/ipfilter/netinet/ip_proxy.h
Directory Properties:
  stable/11/   (props changed)

Modified: stable/10/contrib/ipfilter/opts.h
==
--- stable/10/contrib/ipfilter/opts.h   Wed Aug 31 17:52:11 2016
(r305137)
+++ stable/10/contrib/ipfilter/opts.h   Wed Aug 31 18:00:41 2016
(r305138)
@@ -12,7 +12,11 @@
 #define__OPTS_H__
 
 #ifndefSOLARIS
-#defineSOLARIS (defined(sun) && (defined(__svr4__) || defined(__SVR4)))
+# if defined(sun) && (defined(__svr4__) || defined(__SVR4))
+#  define  SOLARIS 1
+# else
+#  define  SOLARIS 0
+# endif
 #endif
 #defineOPT_REMOVE  0x01
 #defineOPT_DEBUG   0x02

Modified: stable/10/sys/contrib/ipfilter/netinet/ip_compat.h
==
--- stable/10/sys/contrib/ipfilter/netinet/ip_compat.h  Wed Aug 31 17:52:11 
2016(r305137)
+++ stable/10/sys/contrib/ipfilter/netinet/ip_compat.h  Wed Aug 31 18:00:41 
2016(r305138)
@@ -32,7 +32,13 @@
 # define   __KERNEL__
 #endif
 
-#defineSOLARIS (defined(sun) && (defined(__svr4__) || defined(__SVR4)))
+#ifndefSOLARIS
+# if defined(sun) && (defined(__svr4__) || defined(__SVR4))
+#  define  SOLARIS 1
+# else
+#  define  SOLARIS 0
+# endif
+#endif
 
 
 #if defined(__SVR4) || defined(__svr4__) || defined(__sgi)

Modified: stable/10/sys/contrib/ipfilter/netinet/ip_fil.h
==
--- stable/10/sys/contrib/ipfilter/netinet/ip_fil.h Wed Aug 31 17:52:11 
2016(r305137)
+++ stable/10/sys/contrib/ipfilter/netinet/ip_fil.h Wed Aug 31 18:00:41 
2016(r305138)
@@ -29,7 +29,11 @@
 #endif
 
 #ifndefSOLARIS
-# define SOLARIS (defined(sun) && (defined(__svr4__) || defined(__SVR4)))
+# if defined(sun) && (defined(__svr4__) || defined(__SVR4))
+#  define  SOLARIS 1
+# else
+#  define  SOLARIS 0
+# endif
 #endif
 
 #ifndef__P

Modified: stable/10/sys/contrib/ipfilter/netinet/ip_log.c
==
--- stable/10/sys/contrib/ipfilter/netinet/ip_log.c Wed Aug 31 17:52:11 
2016(r305137)
+++ stable/10/sys/contrib/ipfilter/netinet/ip_log.c Wed Aug 31 18:00:41 
2016(r305138)
@@ -19,7 +19,11 @@
 # include 
 #endif
 #ifndef SOLARIS
-# define SOLARIS (defined(sun) && (defined(__svr4__) || defined(__SVR4)))
+# if defined(sun) && (defined(__svr4__) || defined(__SVR4))
+#  define  SOLARIS 1
+# else
+#  define  SOLARIS 0
+# endif
 #endif
 #include 
 #include 

Modified: stable/10/sys/contrib/ipfilter/netinet/ip_nat.h
==
--- stable/10/sys/contrib/ipfilter/netinet/ip_nat.h Wed Aug 31 17:52:11 
2016(r305137)
+++ stable/10/sys/contrib/ipfilter/netinet/ip_nat.h Wed Aug 31 18:00:41 
2016(r305138)
@@ -13,8 +13,12 @@
 #ifndef__IP_NAT_H__
 #define__IP_NAT_H__
 
-#ifndef SOLARIS
-#defineSOLARIS (defined(sun) && (defined(__svr4__) || defined(__SVR4)))
+#ifndefSOLARIS
+# if defined(sun) && (defined(__svr4__) || defined(__SVR4))
+#  define  SOLARIS 1
+# else
+#  define  SOLARIS 0
+# endif
 #endif
 
 #if defined(__STDC__) || defined(__GNUC__) || defined(_AIX51)

Modified: stable/10/sys/contrib/ipfilter/netinet/ip_proxy.h
==
---

svn commit: r305138 - in stable: 10/contrib/ipfilter 10/sys/contrib/ipfilter/netinet 11/contrib/ipfilter 11/sys/contrib/ipfilter/netinet

2016-08-31 Thread Dimitry Andric 
Author: dim
Date: Wed Aug 31 18:00:41 2016
New Revision: 305138
URL: https://svnweb.freebsd.org/changeset/base/305138

Log:
  MFC r304953:
  
  Define ipfilter's SOLARIS macro in a defined and portable way.
  
  Reviewed by:  cy
  Differential Revision: https://reviews.freebsd.org/D7671
  
  MFC r304959 (by kib):
  
  Complete r304953.
  
  Sponsored by: The FreeBSD Foundation
  
  MFC r304964:
  
  Follow-up to r304953, in which I broke the build: apparently the SOLARIS
  macro is defined in lots of different places in ipfilter, so replace all
  of the nonportable definitions with portable ones.
  
  Pointy hat to:dim

Modified:
  stable/11/contrib/ipfilter/opts.h
  stable/11/sys/contrib/ipfilter/netinet/ip_compat.h
  stable/11/sys/contrib/ipfilter/netinet/ip_fil.h
  stable/11/sys/contrib/ipfilter/netinet/ip_log.c
  stable/11/sys/contrib/ipfilter/netinet/ip_nat.h
  stable/11/sys/contrib/ipfilter/netinet/ip_proxy.h
Directory Properties:
  stable/11/   (props changed)

Changes in other areas also in this revision:
Modified:
  stable/10/contrib/ipfilter/opts.h
  stable/10/sys/contrib/ipfilter/netinet/ip_compat.h
  stable/10/sys/contrib/ipfilter/netinet/ip_fil.h
  stable/10/sys/contrib/ipfilter/netinet/ip_log.c
  stable/10/sys/contrib/ipfilter/netinet/ip_nat.h
  stable/10/sys/contrib/ipfilter/netinet/ip_proxy.h
Directory Properties:
  stable/10/   (props changed)

Modified: stable/11/contrib/ipfilter/opts.h
==
--- stable/11/contrib/ipfilter/opts.h   Wed Aug 31 17:52:11 2016
(r305137)
+++ stable/11/contrib/ipfilter/opts.h   Wed Aug 31 18:00:41 2016
(r305138)
@@ -12,7 +12,11 @@
 #define__OPTS_H__
 
 #ifndefSOLARIS
-#defineSOLARIS (defined(sun) && (defined(__svr4__) || defined(__SVR4)))
+# if defined(sun) && (defined(__svr4__) || defined(__SVR4))
+#  define  SOLARIS 1
+# else
+#  define  SOLARIS 0
+# endif
 #endif
 #defineOPT_REMOVE  0x01
 #defineOPT_DEBUG   0x02

Modified: stable/11/sys/contrib/ipfilter/netinet/ip_compat.h
==
--- stable/11/sys/contrib/ipfilter/netinet/ip_compat.h  Wed Aug 31 17:52:11 
2016(r305137)
+++ stable/11/sys/contrib/ipfilter/netinet/ip_compat.h  Wed Aug 31 18:00:41 
2016(r305138)
@@ -32,7 +32,13 @@
 # define   __KERNEL__
 #endif
 
-#defineSOLARIS (defined(sun) && (defined(__svr4__) || defined(__SVR4)))
+#ifndefSOLARIS
+# if defined(sun) && (defined(__svr4__) || defined(__SVR4))
+#  define  SOLARIS 1
+# else
+#  define  SOLARIS 0
+# endif
+#endif
 
 
 #if defined(__SVR4) || defined(__svr4__) || defined(__sgi)

Modified: stable/11/sys/contrib/ipfilter/netinet/ip_fil.h
==
--- stable/11/sys/contrib/ipfilter/netinet/ip_fil.h Wed Aug 31 17:52:11 
2016(r305137)
+++ stable/11/sys/contrib/ipfilter/netinet/ip_fil.h Wed Aug 31 18:00:41 
2016(r305138)
@@ -29,7 +29,11 @@
 #endif
 
 #ifndefSOLARIS
-# define SOLARIS (defined(sun) && (defined(__svr4__) || defined(__SVR4)))
+# if defined(sun) && (defined(__svr4__) || defined(__SVR4))
+#  define  SOLARIS 1
+# else
+#  define  SOLARIS 0
+# endif
 #endif
 
 #ifndef__P

Modified: stable/11/sys/contrib/ipfilter/netinet/ip_log.c
==
--- stable/11/sys/contrib/ipfilter/netinet/ip_log.c Wed Aug 31 17:52:11 
2016(r305137)
+++ stable/11/sys/contrib/ipfilter/netinet/ip_log.c Wed Aug 31 18:00:41 
2016(r305138)
@@ -19,7 +19,11 @@
 # include 
 #endif
 #ifndef SOLARIS
-# define SOLARIS (defined(sun) && (defined(__svr4__) || defined(__SVR4)))
+# if defined(sun) && (defined(__svr4__) || defined(__SVR4))
+#  define  SOLARIS 1
+# else
+#  define  SOLARIS 0
+# endif
 #endif
 #include 
 #include 

Modified: stable/11/sys/contrib/ipfilter/netinet/ip_nat.h
==
--- stable/11/sys/contrib/ipfilter/netinet/ip_nat.h Wed Aug 31 17:52:11 
2016(r305137)
+++ stable/11/sys/contrib/ipfilter/netinet/ip_nat.h Wed Aug 31 18:00:41 
2016(r305138)
@@ -13,8 +13,12 @@
 #ifndef__IP_NAT_H__
 #define__IP_NAT_H__
 
-#ifndef SOLARIS
-#defineSOLARIS (defined(sun) && (defined(__svr4__) || defined(__SVR4)))
+#ifndefSOLARIS
+# if defined(sun) && (defined(__svr4__) || defined(__SVR4))
+#  define  SOLARIS 1
+# else
+#  define  SOLARIS 0
+# endif
 #endif
 
 #if defined(__STDC__) || defined(__GNUC__) || defined(_AIX51)

Modified: stable/11/sys/contrib/ipfilter/netinet/ip_proxy.h
==
---

svn commit: r305137 - head/sys/kern

2016-08-31 Thread Allan Jude 
Author: allanjude
Date: Wed Aug 31 17:52:11 2016
New Revision: 305137
URL: https://svnweb.freebsd.org/changeset/base/305137

Log:
  Eliminate unnecessary loop in _cap_check()
  
  Calling cap_rights_contains() several times with the same inputs is not
  going to produce a different output. The variable being iterated, i, is
  never used inside the for loop.
  
  The loop is actually done in cap_rights_contains()
  
  Submitted by: Ryan Moeller 
  Reviewed by:  oshogbo, ed
  MFC after:1 month
  Differential Revision:https://reviews.freebsd.org/D7369

Modified:
  head/sys/kern/sys_capability.c

Modified: head/sys/kern/sys_capability.c
==
--- head/sys/kern/sys_capability.c  Wed Aug 31 17:36:43 2016
(r305136)
+++ head/sys/kern/sys_capability.c  Wed Aug 31 17:52:11 2016
(r305137)
@@ -150,16 +150,13 @@ static inline int
 _cap_check(const cap_rights_t *havep, const cap_rights_t *needp,
 enum ktr_cap_fail_type type)
 {
-   int i;
 
-   for (i = 0; i < nitems(havep->cr_rights); i++) {
-   if (!cap_rights_contains(havep, needp)) {
+   if (!cap_rights_contains(havep, needp)) {
 #ifdef KTRACE
-   if (KTRPOINT(curthread, KTR_CAPFAIL))
-   ktrcapfail(type, needp, havep);
+   if (KTRPOINT(curthread, KTR_CAPFAIL))
+   ktrcapfail(type, needp, havep);
 #endif
-   return (ENOTCAPABLE);
-   }
+   return (ENOTCAPABLE);
}
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: r305136 - in stable/11/sys: arm64/arm64 arm64/cavium arm64/include dev/pci

2016-08-31 Thread Andrew Turner 
Author: andrew
Date: Wed Aug 31 17:36:43 2016
New Revision: 305136
URL: https://svnweb.freebsd.org/changeset/base/305136

Log:
  MFC 302847, 302848, 302852, 302853:
  Remove the remaining non-INTRNG support from the arm64 code.

Modified:
  stable/11/sys/arm64/arm64/autoconf.c
  stable/11/sys/arm64/arm64/gic_v3.c
  stable/11/sys/arm64/arm64/gic_v3_fdt.c
  stable/11/sys/arm64/arm64/gic_v3_var.h
  stable/11/sys/arm64/arm64/mp_machdep.c
  stable/11/sys/arm64/arm64/nexus.c
  stable/11/sys/arm64/cavium/thunder_pcie_pem_fdt.c
  stable/11/sys/arm64/include/intr.h
  stable/11/sys/dev/pci/pci_host_generic.c
Directory Properties:
  stable/11/   (props changed)

Modified: stable/11/sys/arm64/arm64/autoconf.c
==
--- stable/11/sys/arm64/arm64/autoconf.cWed Aug 31 17:30:21 2016
(r305135)
+++ stable/11/sys/arm64/arm64/autoconf.cWed Aug 31 17:36:43 2016
(r305136)
@@ -81,12 +81,8 @@ static void
 configure_final(void *dummy)
 {
 
-#ifdef INTRNG
/* Enable interrupt reception on this CPU */
intr_enable();
-#else
-   arm_enable_intr();
-#endif
cninit_finish(); 
 
if (bootverbose)

Modified: stable/11/sys/arm64/arm64/gic_v3.c
==
--- stable/11/sys/arm64/arm64/gic_v3.c  Wed Aug 31 17:30:21 2016
(r305135)
+++ stable/11/sys/arm64/arm64/gic_v3.c  Wed Aug 31 17:36:43 2016
(r305136)
@@ -69,7 +69,6 @@ __FBSDID("$FreeBSD$");
 
 static bus_read_ivar_t gic_v3_read_ivar;
 
-#ifdef INTRNG
 static pic_disable_intr_t gic_v3_disable_intr;
 static pic_enable_intr_t gic_v3_enable_intr;
 static pic_map_intr_t gic_v3_map_intr;
@@ -90,18 +89,6 @@ static u_int gic_irq_cpu;
 static u_int sgi_to_ipi[GIC_LAST_SGI - GIC_FIRST_SGI + 1];
 static u_int sgi_first_unused = GIC_FIRST_SGI;
 #endif
-#else
-/* Device and PIC methods */
-static int gic_v3_bind(device_t, u_int, u_int);
-static void gic_v3_dispatch(device_t, struct trapframe *);
-static void gic_v3_eoi(device_t, u_int);
-static void gic_v3_mask_irq(device_t, u_int);
-static void gic_v3_unmask_irq(device_t, u_int);
-#ifdef SMP
-static void gic_v3_init_secondary(device_t);
-static void gic_v3_ipi_send(device_t, cpuset_t, u_int);
-#endif
-#endif
 
 static device_method_t gic_v3_methods[] = {
/* Device interface */
@@ -110,7 +97,6 @@ static device_method_t gic_v3_methods[] 
/* Bus interface */
DEVMETHOD(bus_read_ivar,gic_v3_read_ivar),
 
-#ifdef INTRNG
/* Interrupt controller interface */
DEVMETHOD(pic_disable_intr, gic_v3_disable_intr),
DEVMETHOD(pic_enable_intr,  gic_v3_enable_intr),
@@ -126,18 +112,6 @@ static device_method_t gic_v3_methods[] 
DEVMETHOD(pic_ipi_send, gic_v3_ipi_send),
DEVMETHOD(pic_ipi_setup,gic_v3_ipi_setup),
 #endif
-#else
-   /* PIC interface */
-   DEVMETHOD(pic_bind, gic_v3_bind),
-   DEVMETHOD(pic_dispatch, gic_v3_dispatch),
-   DEVMETHOD(pic_eoi,  gic_v3_eoi),
-   DEVMETHOD(pic_mask, gic_v3_mask_irq),
-   DEVMETHOD(pic_unmask,   gic_v3_unmask_irq),
-#ifdef SMP
-   DEVMETHOD(pic_init_secondary,   gic_v3_init_secondary),
-   DEVMETHOD(pic_ipi_send, gic_v3_ipi_send),
-#endif
-#endif
 
/* End */
DEVMETHOD_END
@@ -188,7 +162,6 @@ static gic_v3_initseq_t gic_v3_secondary
 };
 #endif
 
-#ifdef INTRNG
 uint32_t
 gic_r_read_4(device_t dev, bus_size_t offset)
 {
@@ -224,7 +197,6 @@ gic_r_write_8(device_t dev, bus_size_t o
sc = device_get_softc(dev);
bus_write_8(sc->gic_redists.pcpu[PCPU_GET(cpuid)], offset, val);
 }
-#endif
 
 /*
  * Device interface.
@@ -238,10 +210,8 @@ gic_v3_attach(device_t dev)
int rid;
int err;
size_t i;
-#ifdef INTRNG
u_int irq;
const char *name;
-#endif
 
sc = device_get_softc(dev);
sc->gic_registered = FALSE;
@@ -290,7 +260,6 @@ gic_v3_attach(device_t dev)
if (sc->gic_nirqs > GIC_I_NUM_MAX)
sc->gic_nirqs = GIC_I_NUM_MAX;
 
-#ifdef INTRNG
sc->gic_irqs = malloc(sizeof(*sc->gic_irqs) * sc->gic_nirqs,
M_GIC_V3, M_WAITOK | M_ZERO);
name = device_get_nameunit(dev);
@@ -318,7 +287,6 @@ gic_v3_attach(device_t dev)
return (err);
}
}
-#endif
 
/* Get the number of supported interrupt identifier bits */
sc->gic_idbits = GICD_TYPER_IDBITS(typer);
@@ -334,14 +302,6 @@ gic_v3_attach(device_t dev)
if (err != 0)
return (err);
}
-   /*
-* Full success.
-* Now register PIC to the interrupts handling layer.
-*/
-#ifndef INTRNG
-   arm_register_root_pic(dev, sc->gic_nirqs);
-   sc->gic_registered = TRUE;
-#endif
 
return (0);
 }
@@ -394,7 +354,6 @@

svn commit: r305135 - in stable/11/sys/arm64: arm64 include

2016-08-31 Thread Andrew Turner 
Author: andrew
Date: Wed Aug 31 17:30:21 2016
New Revision: 305135
URL: https://svnweb.freebsd.org/changeset/base/305135

Log:
  MFC 302789:
  Add memmmap on arm64 so we can mmap /dev/mem and /dev/kmem.

Modified:
  stable/11/sys/arm64/arm64/mem.c
  stable/11/sys/arm64/include/memdev.h
Directory Properties:
  stable/11/   (props changed)

Modified: stable/11/sys/arm64/arm64/mem.c
==
--- stable/11/sys/arm64/arm64/mem.c Wed Aug 31 16:56:54 2016
(r305134)
+++ stable/11/sys/arm64/arm64/mem.c Wed Aug 31 17:30:21 2016
(r305135)
@@ -114,3 +114,19 @@ memrw(struct cdev *dev, struct uio *uio,
return (error);
 }
 
+/*
+ * allow user processes to MMAP some memory sections
+ * instead of going through read/write
+ */
+/* ARGSUSED */
+int
+memmmap(struct cdev *dev, vm_ooffset_t offset, vm_paddr_t *paddr,
+int prot __unused, vm_memattr_t *memattr __unused)
+{
+   if (dev2unit(dev) == CDEV_MINOR_MEM)
+   *paddr = offset;
+   else if (dev2unit(dev) == CDEV_MINOR_KMEM)
+   *paddr = vtophys(offset);
+   /* else panic! */
+   return (0);
+}

Modified: stable/11/sys/arm64/include/memdev.h
==
--- stable/11/sys/arm64/include/memdev.hWed Aug 31 16:56:54 2016
(r305134)
+++ stable/11/sys/arm64/include/memdev.hWed Aug 31 17:30:21 2016
(r305135)
@@ -35,6 +35,6 @@
 d_open_t   memopen;
 d_read_t   memrw;
 #definememioctl(d_ioctl_t *)NULL
-#definememmmap (d_mmap_t *)NULL
+d_mmap_t   memmmap;
 
 #endif /* _MACHINE_MEMDEV_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"

svn commit: r305134 - in head/sys/boot/fdt/dts: . arm64

2016-08-31 Thread Jared McNeill 
Author: jmcneill
Date: Wed Aug 31 16:56:54 2016
New Revision: 305134
URL: https://svnweb.freebsd.org/changeset/base/305134

Log:
  Add Allwinner A64 (sun50i) and Pine64 dts files.
  
  Reviewed by:  andrew, imp, manu, wca
  Differential Revision:https://reviews.freebsd.org/D7169

Added:
  head/sys/boot/fdt/dts/arm64/
  head/sys/boot/fdt/dts/arm64/Makefile   (contents, props changed)
  head/sys/boot/fdt/dts/arm64/a64.dtsi   (contents, props changed)
  head/sys/boot/fdt/dts/arm64/pine64_plus.dts   (contents, props changed)
  head/sys/boot/fdt/dts/arm64/sun50i-a64-pine64-common.dtsi   (contents, props 
changed)
  head/sys/boot/fdt/dts/arm64/sun50i-a64-pine64-plus.dts   (contents, props 
changed)
  head/sys/boot/fdt/dts/arm64/sun50i-a64-pine64.dts   (contents, props changed)
  head/sys/boot/fdt/dts/arm64/sun50i-a64.dtsi   (contents, props changed)
Modified:
  head/sys/boot/fdt/dts/Makefile

Modified: head/sys/boot/fdt/dts/Makefile
==
--- head/sys/boot/fdt/dts/Makefile  Wed Aug 31 15:47:58 2016
(r305133)
+++ head/sys/boot/fdt/dts/Makefile  Wed Aug 31 16:56:54 2016
(r305134)
@@ -1,5 +1,5 @@
 # $FreeBSD$
 
-SUBDIR=arm mips powerpc
+SUBDIR=arm arm64 mips powerpc
 
 .include 

Added: head/sys/boot/fdt/dts/arm64/Makefile
==
--- /dev/null   00:00:00 1970   (empty, because file is newly added)
+++ head/sys/boot/fdt/dts/arm64/MakefileWed Aug 31 16:56:54 2016
(r305134)
@@ -0,0 +1,7 @@
+# $FreeBSD$
+
+DTS!=ls *.dts
+
+all: test-dts
+
+.include 

Added: head/sys/boot/fdt/dts/arm64/a64.dtsi
==
--- /dev/null   00:00:00 1970   (empty, because file is newly added)
+++ head/sys/boot/fdt/dts/arm64/a64.dtsiWed Aug 31 16:56:54 2016
(r305134)
@@ -0,0 +1,151 @@
+/*-
+ * Copyright (c) 2016 Jared McNeill 
+ * 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.
+ *
+ * $FreeBSD$
+ */
+
+/ {
+   clocks {
+   pll_hsic: clk@01c20044 {
+   #clock-cells = <0>;
+   compatible = "allwinner,sun50i-a64-pllhsic-clk";
+   reg = <0x01c20044 0x4>;
+   clocks = <>;
+   clock-output-names = "pll_hsic";
+   };
+
+   usb_clk: clk@01c200cc {
+   #clock-cells = <1>;
+   #reset-cells = <1>;
+   compatible = "allwinner,sun8i-a83t-usb-clk";
+   reg = <0x01c200cc 0x4>;
+   clocks = <>, <_hsic>;
+   clock-indices = <8>, <9>,
+   <10>, <11>,
+   <16>, <17>;
+   clock-output-names = "usb_phy0", "usb_phy1",
+"usb_hsic_pll", "usb_hsic_12m",
+"usb_otg_ohci", "usb_ohci0";
+   };
+   };
+
+   soc {
+   watchdog: watchdog@01c20ca0 {
+   compatible = "allwinner,sun6i-a31-wdt";
+   reg = <0x01c20ca0 0x20>;
+   interrupts = ;
+   clocks = <>;
+   };
+
+   nmi_intc: interrupt-controller@01f00c0c {
+   compatible = "allwinner,sun6i-a31-sc-nmi";
+   interrupt-controller;
+   #interrupt-cells = <2>;
+   reg = <0x01f00c0c 0x38>;
+

Re: svn commit: r304510 - head/sys/dev/ixl

2016-08-31 Thread Ed Maste 
On 19 August 2016 at 20:08, Eric Joyner  wrote:
> Author: erj
> Date: Sat Aug 20 00:08:10 2016
> New Revision: 304510
> URL: https://svnweb.freebsd.org/changeset/base/304510
>
> Log:
>   ixlv(4): Fix ixlv(4) not loading when loaded as a kernel module and netmap 
> is enabled.

I did not confirm it's this commit, but kernel builds currently have
this warning:

if_ixlv.o: warning: common of `ixl_crcstrip' overridden by definition
if_ixl.o: warning: defined here
if_ixlv.o: warning: multiple common of `ixl_rx_miss'
if_ixl.o: warning: previous common is here
if_ixlv.o: warning: multiple common of `ixl_rx_miss_bufs'
if_ixl.o: warning: previous common is here
___
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: r305129 - head/sys/vm

2016-08-31 Thread Konstantin Belousov 
On Thu, Sep 01, 2016 at 01:38:31AM +1000, Bruce Evans wrote:
> On Wed, 31 Aug 2016, Konstantin Belousov wrote:
> 
> > Log:
> >  Make swapoff reliable.
> >
> >  The swap_pager_swapoff() function uses trylock for the object lock
> >  before pagein, which means that either i/o to md(4) over swap, or
> >  intensive page faults over swap pager objects might prevent swapoff()
> >  from making any progress. Then the retry < 100 check fails and machine
> >  panics.
> >
> >  If trylock fails, acquire the object lock in the blockable way and
> >  restart the hash bucket walk.  Keep retries logic for now.
> 
> Trylock is difficult to use.  When it fails, there is no way for the
> caller to tell how long it should wait before retrying (much like EAGAIN
> errors for userland).  Sometimes there is deadlock so retrying is worse
> than useless.  Sometimes trylock can detect deadlock, but callers never
> can (otherwise they wouldn't try).
It is actually a way for the caller to tell how long to wait. The lock
should be taken in blockable or sleepable way. Issue is that we have to
drop other locks, to avoid LOR, and restart the algorithm.

This approach, i.e. try fast to get lock with trylock, and do the slow
reset on failure with the required lock already held or not, is very
common pattern.  E.g. it is used in vm_fault() to handle LOR between
vnode lock and vm map lock, it is used in ufs_rename() to avoid LORs
between all vnode locks participating in rename, and in many other
places.

> 
> My version of mtx_trylock_spin_flags() has a timeout in usec.  It returns
> immediately if deadlock is detected.  The timeout is just for convenience
> in simple cases where the caller want to aquire the lock normally but
> doesn't want the unbounded timeout or panic given by
> mtx_trylock_spin().  The important thing is to return error codes like:
> - EDEADLK for deadlock detected (e.g., when the CPU running the thread
>holding the lock is stopped)
> - EMAYBEDEADLK when deadlock is almost detected (e.g., when CPUs are being
>stopped)
> - EAGAIN when no problem is detected but the lock is held
> - ERECURSE when the lock is held but acquiring it recursively would work
>if the caller asked for that
> - ELOR if acquiring the lock would work but give a LOR, and the caller didn't
>ask for LORs.
> 
> The current return value is positive logic for success and doesn't allow
> returning error codes.
Your use of trylock for console locks is very specific.  It is implied by
the intent of console output working in any context, which is not needed
for typical trylock usage.

> 
> In my applications in console drivers, the caller doesn't want the lock if
> it would give a LOR.  Without WITNESS, it is too hard to tell if the lock
> would give a LOR, but some errors are easy to detect.  E.g., acquiring
> sleep lock in a critical section.
Indeed.
___
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: r305133 - head/lib/libc/net

2016-08-31 Thread Andrey A. Chernov 
Author: ache
Date: Wed Aug 31 15:47:58 2016
New Revision: 305133
URL: https://svnweb.freebsd.org/changeset/base/305133

Log:
  Apply the same qsort() usage fix as in r304911 getaddrinfo.c
  qsort() can't be stabilized with just return(-1) alone.
  
  MFC after:  3 days

Modified:
  head/lib/libc/net/name6.c

Modified: head/lib/libc/net/name6.c
==
--- head/lib/libc/net/name6.c   Wed Aug 31 15:32:52 2016(r305132)
+++ head/lib/libc/net/name6.c   Wed Aug 31 15:47:58 2016(r305133)
@@ -185,6 +185,7 @@ struct hp_order {
 #define aio_sa aio_un.aiou_sa
int aio_matchlen;
char *aio_h_addr;
+   int aio_initial_sequence;
 };
 
 static struct   hostent *_hpcopy(struct hostent *, int *);
@@ -711,6 +712,7 @@ _hpreorder(struct hostent *hp)
aio[i].aio_dstscope = gai_addr2scopetype(sa);
aio[i].aio_dstpolicy = match_addrselectpolicy(sa, );
set_source([i], );
+   aio[i].aio_initial_sequence = i;
}
 
/* perform sorting. */
@@ -1045,6 +1047,23 @@ comp_dst(const void *arg1, const void *a
}
 
/* Rule 10: Otherwise, leave the order unchanged. */
+
+   /* 
+* Note that qsort is unstable; so, we can't return zero and 
+* expect the order to be unchanged.
+* That also means we can't depend on the current position of
+* dst2 being after dst1.  We must enforce the initial order
+* with an explicit compare on the original position.
+* The qsort specification requires that "When the same objects 
+* (consisting of width bytes, irrespective of their current 
+* positions in the array) are passed more than once to the 
+* comparison function, the results shall be consistent with one 
+* another."  
+* In other words, If A < B, then we must also return B > A.
+*/
+   if (dst2->aio_initial_sequence < dst1->aio_initial_sequence)
+   return(1);
+
return(-1);
 }
 
___
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: r305129 - head/sys/vm

2016-08-31 Thread Bruce Evans 

On Wed, 31 Aug 2016, Konstantin Belousov wrote:


Log:
 Make swapoff reliable.

 The swap_pager_swapoff() function uses trylock for the object lock
 before pagein, which means that either i/o to md(4) over swap, or
 intensive page faults over swap pager objects might prevent swapoff()
 from making any progress. Then the retry < 100 check fails and machine
 panics.

 If trylock fails, acquire the object lock in the blockable way and
 restart the hash bucket walk.  Keep retries logic for now.


Trylock is difficult to use.  When it fails, there is no way for the
caller to tell how long it should wait before retrying (much like EAGAIN
errors for userland).  Sometimes there is deadlock so retrying is worse
than useless.  Sometimes trylock can detect deadlock, but callers never
can (otherwise they wouldn't try).

My version of mtx_trylock_spin_flags() has a timeout in usec.  It returns
immediately if deadlock is detected.  The timeout is just for convenience
in simple cases where the caller want to aquire the lock normally but
doesn't want the unbounded timeout or panic given by
mtx_trylock_spin().  The important thing is to return error codes like:
- EDEADLK for deadlock detected (e.g., when the CPU running the thread
  holding the lock is stopped)
- EMAYBEDEADLK when deadlock is almost detected (e.g., when CPUs are being
  stopped)
- EAGAIN when no problem is detected but the lock is held
- ERECURSE when the lock is held but acquiring it recursively would work
  if the caller asked for that
- ELOR if acquiring the lock would work but give a LOR, and the caller didn't
  ask for LORs.

The current return value is positive logic for success and doesn't allow
returning error codes.

In my applications in console drivers, the caller doesn't want the lock if
it would give a LOR.  Without WITNESS, it is too hard to tell if the lock
would give a LOR, but some errors are easy to detect.  E.g., acquiring
sleep lock in a critical section.

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: r305132 - head/sys/boot/forth

2016-08-31 Thread Warner Losh 
Author: imp
Date: Wed Aug 31 15:32:52 2016
New Revision: 305132
URL: https://svnweb.freebsd.org/changeset/base/305132

Log:
  Remove accidentally committed stray comment.
  
  Noticed by: dteske@

Modified:
  head/sys/boot/forth/loader.4th

Modified: head/sys/boot/forth/loader.4th
==
--- head/sys/boot/forth/loader.4th  Wed Aug 31 15:06:24 2016
(r305131)
+++ head/sys/boot/forth/loader.4th  Wed Aug 31 15:32:52 2016
(r305132)
@@ -167,8 +167,7 @@ only forth definitions also support-func
 \
 \  Overrides support.4th initialization word with one that does
 \  everything start one does, short of loading the kernel and
-\  modules. Returns a flag. Unsure why we don't just have start
-\  call this...
+\  modules. Returns a flag.
 
 : initialize ( -- flag )
   s" /boot/defaults/loader.conf" initialize
___
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: r297225 - in head/sys: net netinet netinet6

2016-08-31 Thread Slawa Olhovchenkov 
On Thu, Mar 24, 2016 at 07:54:56AM +, George V. Neville-Neil wrote:

> Author: gnn
> Date: Thu Mar 24 07:54:56 2016
> New Revision: 297225
> URL: https://svnweb.freebsd.org/changeset/base/297225
> 
> Log:
>   FreeBSD previously provided route caching for TCP (and UDP). Re-add
>   route caching for TCP, with some improvements. In particular, invalidate
>   the route cache if a new route is added, which might be a better match.
>   The cache is automatically invalidated if the old route is deleted.
>   
>   Submitted by:   Mike Karels
>   Reviewed by:gnn
>   Differential Revision:  https://reviews.freebsd.org/D4306

Can this mergered to stable/10 (in private branch, I mean ABI breakage)?
Have any dependecis of this commit?
___
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: r304436 - in head: . sys/netinet

2016-08-31 Thread Slawa Olhovchenkov 
On Sun, Aug 28, 2016 at 10:20:08AM -0700, Adrian Chadd wrote:

> Hi,
> 
> There are some no brainers here so far(tm):
> 
> working from the bottom up:
> 
> * yeah, the ixgbe locking is a bit silly. Kip's work with iflib and
> converting ixgbe to use that instead of its own locking for managing
> things should remove the bottom two locks
> * the rtalloc1_fib thing - that's odd, because it shouldn't be
> contending there unless there's some temporary redirect that's been
> learnt. What's the routing table look like on your machine? I Remember

As I understund this code:

===
static struct radix_node *
in_matroute(void *v_arg, struct radix_node_head *head)
{
struct radix_node *rn = rn_match(v_arg, head);
struct rtentry *rt = (struct rtentry *)rn;
  
if (rt) {
RT_LOCK(rt);
if (rt->rt_flags & RTPRF_OURS) {
rt->rt_flags &= ~RTPRF_OURS;
rt->rt_expire = 0;
}
RT_UNLOCK(rt);
}
return rn;
}
===

lock congestion will be created for any route: for any route captured
captured exclusive lock (for test).
___
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: r305130 - in head: contrib/elftoolchain/common contrib/elftoolchain/elfcopy contrib/elftoolchain/elfdump contrib/elftoolchain/libelf contrib/elftoolchain/libelftc contrib/elftoolchain/r...

2016-08-31 Thread Ed Maste 
Author: emaste
Date: Wed Aug 31 15:05:04 2016
New Revision: 305130
URL: https://svnweb.freebsd.org/changeset/base/305130

Log:
  Update to ELF Tool Chain r3490
  
  Improvements include:
   * readelf: report all relocation types in rel/rela for MIPS N64
   * readelf: add ELFOSABI_ARM_AEABI
   * elfdump: add ELFOSABI_ARM_AEABI and ELFOSABI_ARM
   * Add recent RISC-V relocations
   * elfcopy: use elftc_timestamp, to support SOURCE_DATE_EPOCH
  
  Sponsored by: The FreeBSD Foundation

Added:
  head/contrib/elftoolchain/libelftc/elftc_timestamp.3
 - copied unchanged from r305126, 
vendor/elftoolchain/dist/libelftc/elftc_timestamp.3
  head/contrib/elftoolchain/libelftc/elftc_timestamp.c
 - copied unchanged from r305126, 
vendor/elftoolchain/dist/libelftc/elftc_timestamp.c
Modified:
  head/contrib/elftoolchain/common/elfdefinitions.h
  head/contrib/elftoolchain/elfcopy/archive.c
  head/contrib/elftoolchain/elfcopy/ascii.c
  head/contrib/elftoolchain/elfcopy/pe.c
  head/contrib/elftoolchain/elfdump/elfdump.c
  head/contrib/elftoolchain/libelf/elf_flagdata.3
  head/contrib/elftoolchain/libelftc/Makefile
  head/contrib/elftoolchain/libelftc/elftc_bfd_find_target.3
  head/contrib/elftoolchain/libelftc/elftc_reloc_type_str.c
  head/contrib/elftoolchain/libelftc/libelftc.h
  head/contrib/elftoolchain/libelftc/libelftc_bfdtarget.c
  head/contrib/elftoolchain/libelftc/libelftc_dem_gnu3.c
  head/contrib/elftoolchain/readelf/readelf.1
  head/contrib/elftoolchain/readelf/readelf.c
  head/lib/libelftc/Makefile
  head/lib/libelftc/elftc_version.c
Directory Properties:
  head/contrib/elftoolchain/   (props changed)
  head/contrib/elftoolchain/elfdump/   (props changed)

Modified: head/contrib/elftoolchain/common/elfdefinitions.h
==
--- head/contrib/elftoolchain/common/elfdefinitions.h   Wed Aug 31 14:49:58 
2016(r305129)
+++ head/contrib/elftoolchain/common/elfdefinitions.h   Wed Aug 31 15:05:04 
2016(r305130)
@@ -23,7 +23,7 @@
  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  * SUCH DAMAGE.
  *
- * $Id: elfdefinitions.h 3455 2016-05-09 13:47:29Z emaste $
+ * $Id: elfdefinitions.h 3485 2016-08-18 13:38:52Z emaste $
  */
 
 /*
@@ -2091,7 +2091,10 @@ _ELF_DEFINE_RELOC(R_RISCV_GNU_VTINHERIT,
 _ELF_DEFINE_RELOC(R_RISCV_GNU_VTENTRY, 42) \
 _ELF_DEFINE_RELOC(R_RISCV_ALIGN,   43) \
 _ELF_DEFINE_RELOC(R_RISCV_RVC_BRANCH,  44) \
-_ELF_DEFINE_RELOC(R_RISCV_RVC_JUMP,45)
+_ELF_DEFINE_RELOC(R_RISCV_RVC_JUMP,45) \
+_ELF_DEFINE_RELOC(R_RISCV_RVC_LUI, 46) \
+_ELF_DEFINE_RELOC(R_RISCV_GPREL_I, 47) \
+_ELF_DEFINE_RELOC(R_RISCV_GPREL_S, 48)
 
 #define_ELF_DEFINE_SPARC_RELOCATIONS() \
 _ELF_DEFINE_RELOC(R_SPARC_NONE,0)  \

Modified: head/contrib/elftoolchain/elfcopy/archive.c
==
--- head/contrib/elftoolchain/elfcopy/archive.c Wed Aug 31 14:49:58 2016
(r305129)
+++ head/contrib/elftoolchain/elfcopy/archive.c Wed Aug 31 15:05:04 2016
(r305130)
@@ -38,7 +38,7 @@
 
 #include "elfcopy.h"
 
-ELFTC_VCSID("$Id: archive.c 3287 2015-12-31 16:58:48Z emaste $");
+ELFTC_VCSID("$Id: archive.c 3490 2016-08-31 00:12:22Z emaste $");
 
 #define _ARMAG_LEN 8   /* length of ar magic string */
 #define _ARHDR_LEN 60  /* length of ar header */
@@ -440,6 +440,7 @@ ac_write_objs(struct elfcopy *ecp, int o
struct archive  *a;
struct archive_entry*entry;
struct ar_obj   *obj;
+   time_t   timestamp;
int  nr;
 
if ((a = archive_write_new()) == NULL)
@@ -450,7 +451,9 @@ ac_write_objs(struct elfcopy *ecp, int o
/* Write the archive symbol table, even if it's empty. */
entry = archive_entry_new();
archive_entry_copy_pathname(entry, "/");
-   archive_entry_set_mtime(entry, time(NULL), 0);
+   if (elftc_timestamp() != 0)
+   err(EXIT_FAILURE, "elftc_timestamp");
+   archive_entry_set_mtime(entry, timestamp, 0);
archive_entry_set_size(entry, (ecp->s_cnt + 1) * sizeof(uint32_t) +
ecp->s_sn_sz);
AC(archive_write_header(a, entry));

Modified: head/contrib/elftoolchain/elfcopy/ascii.c
==
--- head/contrib/elftoolchain/elfcopy/ascii.c   Wed Aug 31 14:49:58 2016
(r305129)
+++ head/contrib/elftoolchain/elfcopy/ascii.c   Wed Aug 31 15:05:04 2016
(r305130)
@@ -36,7 +36,7 @@
 
 #include "elfcopy.h"
 
-ELFTC_VCSID("$Id: ascii.c 3446 2016-05-03 01:31:17Z emaste $");
+ELFTC_VCSID("$Id: ascii.c 3487 2016-08-24 18:12:08Z emaste $");
 
 static void append_data(struct section *s, const void *buf, size_t sz);
 static char hex_digit(uint8_t n);
@@ -251,7

svn commit: r305129 - head/sys/vm

2016-08-31 Thread Konstantin Belousov 
Author: kib
Date: Wed Aug 31 14:49:58 2016
New Revision: 305129
URL: https://svnweb.freebsd.org/changeset/base/305129

Log:
  Make swapoff reliable.
  
  The swap_pager_swapoff() function uses trylock for the object lock
  before pagein, which means that either i/o to md(4) over swap, or
  intensive page faults over swap pager objects might prevent swapoff()
  from making any progress. Then the retry < 100 check fails and machine
  panics.
  
  If trylock fails, acquire the object lock in the blockable way and
  restart the hash bucket walk.  Keep retries logic for now.
  
  Reported and tested by:   pho
  Reviewed by:  alc, markj
  Sponsored by: The FreeBSD Foundation
  MFC after:2 weeks
  Differential revision:https://reviews.freebsd.org/D7688

Modified:
  head/sys/vm/swap_pager.c

Modified: head/sys/vm/swap_pager.c
==
--- head/sys/vm/swap_pager.cWed Aug 31 13:58:32 2016(r305128)
+++ head/sys/vm/swap_pager.cWed Aug 31 14:49:58 2016(r305129)
@@ -1688,36 +1688,49 @@ static void
 swap_pager_swapoff(struct swdevt *sp)
 {
struct swblock *swap;
+   vm_object_t locked_obj, object;
+   vm_pindex_t pindex;
int i, j, retries;
 
sx_assert(_syscall_lock, SA_XLOCKED);
 
retries = 0;
+   locked_obj = NULL;
 full_rescan:
mtx_lock(_mtx);
for (i = 0; i <= swhash_mask; i++) { /* '<=' is correct here */
 restart:
for (swap = swhash[i]; swap != NULL; swap = swap->swb_hnext) {
-   vm_object_t object = swap->swb_object;
-   vm_pindex_t pindex = swap->swb_index;
+   object = swap->swb_object;
+   pindex = swap->swb_index;
for (j = 0; j < SWAP_META_PAGES; ++j) {
-   if (swp_pager_isondev(swap->swb_pages[j], sp)) {
-   /* avoid deadlock */
+   if (!swp_pager_isondev(swap->swb_pages[j], sp))
+   continue;
+   if (locked_obj != object) {
+   if (locked_obj != NULL)
+   VM_OBJECT_WUNLOCK(locked_obj);
+   locked_obj = object;
if (!VM_OBJECT_TRYWLOCK(object)) {
-   break;
-   } else {
mtx_unlock(_mtx);
-   swp_pager_force_pagein(object,
-   pindex + j);
-   VM_OBJECT_WUNLOCK(object);
+   /* Depends on type-stability. */
+   VM_OBJECT_WLOCK(object);
mtx_lock(_mtx);
goto restart;
}
}
+   MPASS(locked_obj == object);
+   mtx_unlock(_mtx);
+   swp_pager_force_pagein(object, pindex + j);
+   mtx_lock(_mtx);
+   goto restart;
}
}
}
mtx_unlock(_mtx);
+   if (locked_obj != NULL) {
+   VM_OBJECT_WUNLOCK(locked_obj);
+   locked_obj = NULL;
+   }
if (sp->sw_used) {
/*
 * Objects may be locked or paging to the device being
___
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: r305107 - head/sys/boot/forth

2016-08-31 Thread Warner Losh 
On Wed, Aug 31, 2016 at 8:41 AM, Devin Teske  wrote:
> Please remove the following:
>
> "Unsure why we don't just have start"
>
> Others beside yourself have great understanding of the Forth and we don't 
> need questioning comments in the code.

To be fair, it was just left in there accidentally. However, could you
explain it to me? That answer is insulting without being actually
helpful in resolving the situation.

Warner
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"

Re: svn commit: r305107 - head/sys/boot/forth

2016-08-31 Thread Devin Teske 
Please remove the following:

"Unsure why we don't just have start"

Others beside yourself have great understanding of the Forth and we don't need 
questioning comments in the code.
-- 
Devin

> On Aug 30, 2016, at 8:55 PM, Warner Losh  wrote:
> 
> Unsure why we don't just have start

___
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: r305128 - head/sys/arm64/arm64

2016-08-31 Thread Andrew Turner 
Author: andrew
Date: Wed Aug 31 13:58:32 2016
New Revision: 305128
URL: https://svnweb.freebsd.org/changeset/base/305128

Log:
  Also handle instruction traps. We might hit these when the page we are
  executing is being promoted to a superpage.
  
  Obtained from:ABT Systems Ltd
  MFC after:1 month
  Sponsored by: The FreeBSD Foundation

Modified:
  head/sys/arm64/arm64/trap.c

Modified: head/sys/arm64/arm64/trap.c
==
--- head/sys/arm64/arm64/trap.c Wed Aug 31 13:51:41 2016(r305127)
+++ head/sys/arm64/arm64/trap.c Wed Aug 31 13:58:32 2016(r305128)
@@ -285,6 +285,7 @@ do_el1h_sync(struct trapframe *frame)
print_registers(frame);
printf(" esr: %.8lx\n", esr);
panic("VFP exception in the kernel");
+   case EXCP_INSN_ABORT:
case EXCP_DATA_ABORT:
far = READ_SPECIALREG(far_el1);
intr_enable();
___
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: r305127 - vendor/elftoolchain/elftoolchain-r3490

2016-08-31 Thread Ed Maste 
Author: emaste
Date: Wed Aug 31 13:51:41 2016
New Revision: 305127
URL: https://svnweb.freebsd.org/changeset/base/305127

Log:
  Tag ELF Tool Chain r3490

Added:
  vendor/elftoolchain/elftoolchain-r3490/
 - copied from r305126, vendor/elftoolchain/dist/
___
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: r305126 - in vendor/elftoolchain/dist: common elfcopy elfdump libelf libelftc readelf

2016-08-31 Thread Ed Maste 
Author: emaste
Date: Wed Aug 31 13:50:23 2016
New Revision: 305126
URL: https://svnweb.freebsd.org/changeset/base/305126

Log:
  Import ELF Tool Chain snapshot at revision 3490
  
  From http://svn.code.sf.net/p/elftoolchain/code/

Added:
  vendor/elftoolchain/dist/libelftc/elftc_timestamp.3   (contents, props 
changed)
  vendor/elftoolchain/dist/libelftc/elftc_timestamp.c   (contents, props 
changed)
Modified:
  vendor/elftoolchain/dist/common/elfdefinitions.h
  vendor/elftoolchain/dist/elfcopy/archive.c
  vendor/elftoolchain/dist/elfcopy/ascii.c
  vendor/elftoolchain/dist/elfcopy/pe.c
  vendor/elftoolchain/dist/elfdump/elfdump.c
  vendor/elftoolchain/dist/libelf/elf_flagdata.3
  vendor/elftoolchain/dist/libelftc/Makefile
  vendor/elftoolchain/dist/libelftc/elftc_bfd_find_target.3
  vendor/elftoolchain/dist/libelftc/elftc_reloc_type_str.c
  vendor/elftoolchain/dist/libelftc/libelftc.h
  vendor/elftoolchain/dist/libelftc/libelftc_bfdtarget.c
  vendor/elftoolchain/dist/libelftc/libelftc_dem_gnu3.c
  vendor/elftoolchain/dist/readelf/readelf.1
  vendor/elftoolchain/dist/readelf/readelf.c

Modified: vendor/elftoolchain/dist/common/elfdefinitions.h
==
--- vendor/elftoolchain/dist/common/elfdefinitions.hWed Aug 31 13:16:40 
2016(r305125)
+++ vendor/elftoolchain/dist/common/elfdefinitions.hWed Aug 31 13:50:23 
2016(r305126)
@@ -23,7 +23,7 @@
  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  * SUCH DAMAGE.
  *
- * $Id: elfdefinitions.h 3455 2016-05-09 13:47:29Z emaste $
+ * $Id: elfdefinitions.h 3485 2016-08-18 13:38:52Z emaste $
  */
 
 /*
@@ -2091,7 +2091,10 @@ _ELF_DEFINE_RELOC(R_RISCV_GNU_VTINHERIT,
 _ELF_DEFINE_RELOC(R_RISCV_GNU_VTENTRY, 42) \
 _ELF_DEFINE_RELOC(R_RISCV_ALIGN,   43) \
 _ELF_DEFINE_RELOC(R_RISCV_RVC_BRANCH,  44) \
-_ELF_DEFINE_RELOC(R_RISCV_RVC_JUMP,45)
+_ELF_DEFINE_RELOC(R_RISCV_RVC_JUMP,45) \
+_ELF_DEFINE_RELOC(R_RISCV_RVC_LUI, 46) \
+_ELF_DEFINE_RELOC(R_RISCV_GPREL_I, 47) \
+_ELF_DEFINE_RELOC(R_RISCV_GPREL_S, 48)
 
 #define_ELF_DEFINE_SPARC_RELOCATIONS() \
 _ELF_DEFINE_RELOC(R_SPARC_NONE,0)  \

Modified: vendor/elftoolchain/dist/elfcopy/archive.c
==
--- vendor/elftoolchain/dist/elfcopy/archive.c  Wed Aug 31 13:16:40 2016
(r305125)
+++ vendor/elftoolchain/dist/elfcopy/archive.c  Wed Aug 31 13:50:23 2016
(r305126)
@@ -38,7 +38,7 @@
 
 #include "elfcopy.h"
 
-ELFTC_VCSID("$Id: archive.c 3287 2015-12-31 16:58:48Z emaste $");
+ELFTC_VCSID("$Id: archive.c 3490 2016-08-31 00:12:22Z emaste $");
 
 #define _ARMAG_LEN 8   /* length of ar magic string */
 #define _ARHDR_LEN 60  /* length of ar header */
@@ -440,6 +440,7 @@ ac_write_objs(struct elfcopy *ecp, int o
struct archive  *a;
struct archive_entry*entry;
struct ar_obj   *obj;
+   time_t   timestamp;
int  nr;
 
if ((a = archive_write_new()) == NULL)
@@ -450,7 +451,9 @@ ac_write_objs(struct elfcopy *ecp, int o
/* Write the archive symbol table, even if it's empty. */
entry = archive_entry_new();
archive_entry_copy_pathname(entry, "/");
-   archive_entry_set_mtime(entry, time(NULL), 0);
+   if (elftc_timestamp() != 0)
+   err(EXIT_FAILURE, "elftc_timestamp");
+   archive_entry_set_mtime(entry, timestamp, 0);
archive_entry_set_size(entry, (ecp->s_cnt + 1) * sizeof(uint32_t) +
ecp->s_sn_sz);
AC(archive_write_header(a, entry));

Modified: vendor/elftoolchain/dist/elfcopy/ascii.c
==
--- vendor/elftoolchain/dist/elfcopy/ascii.cWed Aug 31 13:16:40 2016
(r305125)
+++ vendor/elftoolchain/dist/elfcopy/ascii.cWed Aug 31 13:50:23 2016
(r305126)
@@ -36,7 +36,7 @@
 
 #include "elfcopy.h"
 
-ELFTC_VCSID("$Id: ascii.c 3446 2016-05-03 01:31:17Z emaste $");
+ELFTC_VCSID("$Id: ascii.c 3487 2016-08-24 18:12:08Z emaste $");
 
 static void append_data(struct section *s, const void *buf, size_t sz);
 static char hex_digit(uint8_t n);
@@ -251,6 +251,7 @@ create_elf_from_srec(struct elfcopy *ecp
sec_index = 1;
sec_addr = entry = 0;
while (fgets(line, _LINE_BUFSZ, ifp) != NULL) {
+   sz = 0;
if (line[0] == '\r' || line[0] == '\n')
continue;
if (line[0] == '$' && line[1] == '$') {

Modified: vendor/elftoolchain/dist/elfcopy/pe.c
==
--- vendor/elftoolchain/dist/elfcopy/pe.c   Wed Aug 31 13:16:40 2016
(r305125)
+++ vendor/elftoolchain/dist/elfcopy/pe.c

Re: svn commit: r304436 - in head: . sys/netinet

2016-08-31 Thread Slawa Olhovchenkov 
On Sun, Aug 28, 2016 at 10:20:08AM -0700, Adrian Chadd wrote:

> Then the rest of the big entries are just a combination of rtentry
> locking, tcp timer locking, zfs locking and madvise locking. There's
> some sowakeup locking there as well, from the socket producer/consumer
> locking.

On some high network speed I am see like kevent's congestion:

22.79%  [89522]__mtx_lock_sleep @ /boot/kernel.VSTREAM/kernel
 82.99%  [74295] knote
  99.97%  [74269]  sowakeup
   99.07%  [73581]   tcp_do_segment
100.0%  [73581]tcp_input
 100.0%  [73581] ip_input
  100.0%  [73581]  netisr_dispatch_src
   100.0%  [73581]   ether_demux
100.0%  [73581]ether_nh_input
 100.0%  [73581] netisr_dispatch_src
  68.88%  [50683]  t4_eth_rx @ 
/boot/kernel.VSTREAM/if_cxgbe.ko
   100.0%  [50683]   service_iq
100.0%  [50683]t4_intr
 100.0%  [50683] intr_event_execute_handlers @ 
/boot/kernel.VSTREAM/kernel
  100.0%  [50683]  ithread_loop
   100.0%  [50683]   fork_exit
  31.12%  [22898]  tcp_lro_flush
   97.30%  [22280]   service_iq @ 
/boot/kernel.VSTREAM/if_cxgbe.ko
100.0%  [22280]t4_intr
 100.0%  [22280] intr_event_execute_handlers @ 
/boot/kernel.VSTREAM/kernel
  100.0%  [22280]  ithread_loop
   100.0%  [22280]   fork_exit
   01.55%  [354] tcp_lro_flush_inactive
100.0%  [354]  service_iq @ 
/boot/kernel.VSTREAM/if_cxgbe.ko
 100.0%  [354]   t4_intr
  100.0%  [354]intr_event_execute_handlers @ 
/boot/kernel.VSTREAM/kernel
   100.0%  [354] ithread_loop
   01.15%  [264] tcp_lro_rx
100.0%  [264]  t4_eth_rx @ 
/boot/kernel.VSTREAM/if_cxgbe.ko
 100.0%  [264]   service_iq
  100.0%  [264]t4_intr
   100.0%  [264] intr_event_execute_handlers @ 
/boot/kernel.VSTREAM/kernel

Can I do some recomendation?
___
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: r305125 - in head: lib/libstand sys/boot/common sys/boot/efi/libefi sys/boot/efi/loader sys/boot/i386/libi386 sys/boot/i386/loader sys/boot/pc98/loader

2016-08-31 Thread Baptiste Daroussin 
On Wed, Aug 31, 2016 at 01:16:40PM +, Baptiste Daroussin wrote:
> Author: bapt
> Date: Wed Aug 31 13:16:40 2016
> New Revision: 305125
> URL: https://svnweb.freebsd.org/changeset/base/305125
> 
> Log:
>   Netboot: allow both tftpfs and nfs in both pxeboot and loader.efi
>   
>   Add a new 'netproto' variable which can be set for now to
>   NET_TFTP or NET_NFS (default to NET_NONE)
>   
>   From the dhcp options if one sets the root-path option to:
>   "ip:path", the loader will consider it is booting over NFS
>   (meaning same behaviour as the default current behaviour)
>   
>   if the dhcp option "tftp server address" is set (option 150)
>   the loader will consider it is booting over tftpfs, it will then
>   consider the root-path options with 2 possible case
>   1. "path" then the IP of the tftp server will be the one passed by
>   the option 150, and the files will be retrieved under "path" on the tftp
>   server
>   2. "ip:path" then the IP of the tftp server will be the one passed in
>   the option "overwritting the IP from the option 150.
>   
>   We could not "abuse" the rootpath option in the form or tftp://ip:path 
> because
>   this is already used for other purpose by iPXE preventing any chainload from
>   iPXE to the FreeBSD loader.
>   
>   Given at each open(), the loader loops over all available filesystems and 
> keep
>   the "best" error, we needed to prevent tftpfs to fallback on nfs and vice 
> versa.
>   the tftpfs and nfs implementation in libstand now return EINVAL early if
>   'netproto' for that purpose.
>   
>   Reviewed by:tsoome
>   Relnotes:   yes
>   Sponsored by:   Gandi.net
>   Differential Revision:  https://reviews.freebsd.org/D7628

I forgot to mention a guard was added to libstand's net.h, because it is
included multiple times for some loaders

Best regards,
Bapt


signature.asc
Description: PGP signature

svn commit: r305125 - in head: lib/libstand sys/boot/common sys/boot/efi/libefi sys/boot/efi/loader sys/boot/i386/libi386 sys/boot/i386/loader sys/boot/pc98/loader

2016-08-31 Thread Baptiste Daroussin 
Author: bapt
Date: Wed Aug 31 13:16:40 2016
New Revision: 305125
URL: https://svnweb.freebsd.org/changeset/base/305125

Log:
  Netboot: allow both tftpfs and nfs in both pxeboot and loader.efi
  
  Add a new 'netproto' variable which can be set for now to
  NET_TFTP or NET_NFS (default to NET_NONE)
  
  From the dhcp options if one sets the root-path option to:
  "ip:path", the loader will consider it is booting over NFS
  (meaning same behaviour as the default current behaviour)
  
  if the dhcp option "tftp server address" is set (option 150)
  the loader will consider it is booting over tftpfs, it will then
  consider the root-path options with 2 possible case
  1. "path" then the IP of the tftp server will be the one passed by
  the option 150, and the files will be retrieved under "path" on the tftp
  server
  2. "ip:path" then the IP of the tftp server will be the one passed in
  the option "overwritting the IP from the option 150.
  
  We could not "abuse" the rootpath option in the form or tftp://ip:path because
  this is already used for other purpose by iPXE preventing any chainload from
  iPXE to the FreeBSD loader.
  
  Given at each open(), the loader loops over all available filesystems and keep
  the "best" error, we needed to prevent tftpfs to fallback on nfs and vice 
versa.
  the tftpfs and nfs implementation in libstand now return EINVAL early if
  'netproto' for that purpose.
  
  Reviewed by:  tsoome
  Relnotes: yes
  Sponsored by: Gandi.net
  Differential Revision:https://reviews.freebsd.org/D7628

Modified:
  head/lib/libstand/bootp.c
  head/lib/libstand/bootp.h
  head/lib/libstand/globals.c
  head/lib/libstand/net.h
  head/lib/libstand/nfs.c
  head/lib/libstand/tftp.c
  head/sys/boot/common/dev_net.c
  head/sys/boot/efi/libefi/Makefile
  head/sys/boot/efi/loader/Makefile
  head/sys/boot/efi/loader/conf.c
  head/sys/boot/i386/libi386/pxe.c
  head/sys/boot/i386/loader/conf.c
  head/sys/boot/pc98/loader/conf.c

Modified: head/lib/libstand/bootp.c
==
--- head/lib/libstand/bootp.c   Wed Aug 31 12:29:04 2016(r305124)
+++ head/lib/libstand/bootp.c   Wed Aug 31 13:16:40 2016(r305125)
@@ -411,6 +411,10 @@ vend_rfc1048(cp, len)
bcopy(cp, _serverip.s_addr,
  sizeof(dhcp_serverip.s_addr));
}
+   if (tag == TAG_TFTP_SERVER) {
+   bcopy(cp, _addr,
+ sizeof(tftpip.s_addr));
+   }
 #endif
cp += size;
}

Modified: head/lib/libstand/bootp.h
==
--- head/lib/libstand/bootp.h   Wed Aug 31 12:29:04 2016(r305124)
+++ head/lib/libstand/bootp.h   Wed Aug 31 13:16:40 2016(r305125)
@@ -106,6 +106,7 @@ struct bootp {
 #define TAG_T2 ((unsigned char)  59)
 #define TAG_CLASSID((unsigned char)  60)
 #define TAG_CLIENTID   ((unsigned char)  61)
+#define TAG_TFTP_SERVER((unsigned char) 150)
 #endif
 
 #define TAG_END((unsigned char) 255)

Modified: head/lib/libstand/globals.c
==
--- head/lib/libstand/globals.c Wed Aug 31 12:29:04 2016(r305124)
+++ head/lib/libstand/globals.c Wed Aug 31 13:16:40 2016(r305125)
@@ -25,12 +25,14 @@ charhostname[FNAME_SIZE];   /* our hostn
 inthostnamelen;
 char   domainname[FNAME_SIZE]; /* our DNS domain */
 intdomainnamelen;
+intnetproto = NET_NONE;/* Network prototol */
 char   ifname[IFNAME_SIZE];/* name of interface (e.g. "le0") */
 struct in_addr myip;   /* my ip address */
 struct in_addr nameip; /* DNS server ip address */
 struct in_addr rootip; /* root ip address */
 struct in_addr swapip; /* swap ip address */
 struct in_addr gateip; /* gateway ip address */
+struct in_addr tftpip; /* TFTP ip address */
 n_long netmask = 0xff00;   /* subnet or net mask */
 u_int  intf_mtu;   /* interface mtu from bootp/dhcp */
 interrno;  /* our old friend */

Modified: head/lib/libstand/net.h
==
--- head/lib/libstand/net.h Wed Aug 31 12:29:04 2016(r305124)
+++ head/lib/libstand/net.h Wed Aug 31 13:16:40 2016(r305125)
@@ -36,6 +36,8 @@
  * $FreeBSD$
  */
 
+#ifndef _STAND_NET_H
+#define _STAND_NET_H
 #ifndef _KERNEL/* XXX - see  */
 #undef __IPADDR
 #define __IPADDR(x)htonl((u_int32_t)(x))
@@ -45,6 +47,12 @@
 
 #define BA { 0xff, 0xff, 0xff, 0xff, 0xff, 0xff }
 
+enum net_proto {
+   NET_NONE,
+   NET_NFS,
+   NET_TFTP
+};
+
 /* Returns

svn commit: r305124 - head/sys/sys

2016-08-31 Thread Mateusz Guzik 
Author: mjg
Date: Wed Aug 31 12:29:04 2016
New Revision: 305124
URL: https://svnweb.freebsd.org/changeset/base/305124

Log:
  fd: effectively revert r305091
  
  Turns out fd_lastfile can survive being -1 for some processes, giving
  incorrect results with the cast.
  
  Noted by: cem

Modified:
  head/sys/sys/filedesc.h

Modified: head/sys/sys/filedesc.h
==
--- head/sys/sys/filedesc.h Wed Aug 31 11:55:31 2016(r305123)
+++ head/sys/sys/filedesc.h Wed Aug 31 12:29:04 2016(r305124)
@@ -201,7 +201,7 @@ fget_locked(struct filedesc *fdp, int fd
 
FILEDESC_LOCK_ASSERT(fdp);
 
-   if ((u_int)fd > fdp->fd_lastfile)
+   if (fd < 0 || fd > fdp->fd_lastfile)
return (NULL);
 
return (fdp->fd_ofiles[fd].fde_file);
@@ -213,7 +213,7 @@ fdeget_locked(struct filedesc *fdp, int 
 
FILEDESC_LOCK_ASSERT(fdp);
 
-   if ((u_int)fd > fdp->fd_lastfile)
+   if (fd < 0 || fd > fdp->fd_lastfile)
return (NULL);
 
return (>fd_ofiles[fd]);
___
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: r305123 - head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs

2016-08-31 Thread Alexander Motin 
Author: mav
Date: Wed Aug 31 11:55:31 2016
New Revision: 305123
URL: https://svnweb.freebsd.org/changeset/base/305123

Log:
  Fix kernel panic when inheriting properties without default.
  
  There are two writable hidden properties "iscsioptions" and "stmf_sbd_lu",
  that have no default string value.  Attempt to unset them or replicate
  caused kernel panic.  This simple bandaid seems fixes the problem nicely.
  
  MFC after:2 weeks

Modified:
  head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dsl_prop.c

Modified: head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dsl_prop.c
==
--- head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dsl_prop.c  Wed Aug 
31 11:51:52 2016(r305122)
+++ head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dsl_prop.c  Wed Aug 
31 11:55:31 2016(r305123)
@@ -54,6 +54,8 @@ dodefault(zfs_prop_t prop, int intsz, in
return (SET_ERROR(ENOENT));
 
if (zfs_prop_get_type(prop) == PROP_TYPE_STRING) {
+   if (zfs_prop_default_string(prop) == NULL)
+   return (SET_ERROR(ENOENT));
if (intsz != 1)
return (SET_ERROR(EOVERFLOW));
(void) strncpy(buf, zfs_prop_default_string(prop),
___
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: r305122 - head/sys/netipsec

2016-08-31 Thread Andrey V. Elsukov 
Author: ae
Date: Wed Aug 31 11:51:52 2016
New Revision: 305122
URL: https://svnweb.freebsd.org/changeset/base/305122

Log:
  Remove redundant sanity checks from ipsec[46]_common_input_cb().
  
  This check already has been done in the each protocol callback.

Modified:
  head/sys/netipsec/ipsec_input.c

Modified: head/sys/netipsec/ipsec_input.c
==
--- head/sys/netipsec/ipsec_input.c Wed Aug 31 11:10:39 2016
(r305121)
+++ head/sys/netipsec/ipsec_input.c Wed Aug 31 11:51:52 2016
(r305122)
@@ -334,14 +334,6 @@ ipsec4_common_input_cb(struct mbuf *m, s
sproto == IPPROTO_IPCOMP,
("unexpected security protocol %u", sproto));
 
-   /* Sanity check */
-   if (m == NULL) {
-   DPRINTF(("%s: null mbuf", __func__));
-   IPSEC_ISTAT(sproto, badkcr);
-   KEY_FREESAV();
-   return EINVAL;
-   }
-
if (skip != 0) {
/*
 * Fix IPv4 header
@@ -615,14 +607,6 @@ ipsec6_common_input_cb(struct mbuf *m, s
sproto == IPPROTO_IPCOMP,
("unexpected security protocol %u", sproto));
 
-   /* Sanity check */
-   if (m == NULL) {
-   DPRINTF(("%s: null mbuf", __func__));
-   IPSEC_ISTAT(sproto, badkcr);
-   error = EINVAL;
-   goto bad;
-   }
-
/* Fix IPv6 header */
if (m->m_len < sizeof(struct ip6_hdr) &&
(m = m_pullup(m, sizeof(struct ip6_hdr))) == NULL) {
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"

svn commit: r305121 - head/sys/dev/syscons

2016-08-31 Thread Bruce Evans 
Author: bde
Date: Wed Aug 31 11:10:39 2016
New Revision: 305121
URL: https://svnweb.freebsd.org/changeset/base/305121

Log:
  Add some locking to sc_cngetc().
  
  Keyboard input needs Giant locking, and that is not possible to do
  correctly here.  Use mtx_trylock() and proceed unlocked as before if
  we can't acquire Giant (non-recursively), except in kdb mode don't
  even try to acquire Giant.  Everything here is a hack, but it often
  works.  Even if mtx_trylock() succeeds, this might be a LOR.
  
  Keyboard input also needs screen locking, to handle screen updates
  and switches.  Add this, using the same simplistic screen locking
  as for sc_cnputc().
  
  Giant must be acquired before the screen lock, and the screen lock
  must be dropped when calling the keyboard driver (else it would get a
  harmless LOR if it tries to acquire Giant).  It was intended that sc
  cn open/close hide the locking calls, and they do for i/o functions
  functions except for this complication.
  
  Non-console keyboard input is still only Giant-locked, with screen
  locking in some called functions.  This is correct for the keyboard
  parts only.
  
  When Giant cannot be acquired properly, atkbd and kbdmux tend to race
  and work (they assume that the caller acquired Giant properly and don't
  try to acquire it again or check that it has been acquired, and the
  races rarely matter), while ukbd tends to deadlock or panic (since it
  does the opposite, and has other usb threads to deadlock with).
  
  The keyboard (Giant) locking here does very little, but the screen
  locking completes screen locking for console mode except for not
  detecting or handling deadlock.

Modified:
  head/sys/dev/syscons/syscons.c
  head/sys/dev/syscons/syscons.h

Modified: head/sys/dev/syscons/syscons.c
==
--- head/sys/dev/syscons/syscons.c  Wed Aug 31 10:45:53 2016
(r305120)
+++ head/sys/dev/syscons/syscons.c  Wed Aug 31 11:10:39 2016
(r305121)
@@ -1649,11 +1649,33 @@ sc_cnterm(struct consdev *cp)
 
 static void sccnclose(sc_softc_t *sc, struct sc_cnstate *sp);
 static int sc_cngetc_locked(struct sc_cnstate *sp);
+static void sccnkbdlock(sc_softc_t *sc, struct sc_cnstate *sp);
+static void sccnkbdunlock(sc_softc_t *sc, struct sc_cnstate *sp);
 static void sccnopen(sc_softc_t *sc, struct sc_cnstate *sp, int flags);
 static void sccnscrlock(sc_softc_t *sc, struct sc_cnstate *sp);
 static void sccnscrunlock(sc_softc_t *sc, struct sc_cnstate *sp);
 
 static void
+sccnkbdlock(sc_softc_t *sc, struct sc_cnstate *sp)
+{
+/*
+ * Locking method: hope for the best.
+ * The keyboard is supposed to be Giant locked.  We can't handle that
+ * in general.  The kdb_active case here is not safe, and we will
+ * proceed without the lock in all cases.
+ */
+sp->kbd_locked = !kdb_active && mtx_trylock();
+}
+
+static void
+sccnkbdunlock(sc_softc_t *sc, struct sc_cnstate *sp)
+{
+if (sp->kbd_locked)
+   mtx_unlock();
+sp->kbd_locked = FALSE;
+}
+
+static void
 sccnscrlock(sc_softc_t *sc, struct sc_cnstate *sp)
 {
 SC_VIDEO_LOCK(sc);
@@ -1674,11 +1696,14 @@ sccnopen(sc_softc_t *sc, struct sc_cnsta
 
 sp->kbd_opened = FALSE;
 sp->scr_opened = FALSE;
+sp->kbd_locked = FALSE;
 
 /* Opening the keyboard is optional. */
 if (!(flags & 1) || sc->kbd == NULL)
goto over_keyboard;
 
+sccnkbdlock(sc, sp);
+
 /*
  * Make sure the keyboard is accessible even when the kbd device
  * driver is disabled.
@@ -1726,6 +1751,7 @@ sccnclose(sc_softc_t *sc, struct sc_cnst
 
 kbdd_disable(sc->kbd);
 sp->kbd_opened = FALSE;
+sccnkbdunlock(sc, sp);
 }
 
 /*
@@ -1751,6 +1777,7 @@ sc_cngrab(struct consdev *cp)
 if (lev >= 0 && lev < 2) {
sccnopen(sc, >grab_state[lev], 1 | 2);
sccnscrunlock(sc, >grab_state[lev]);
+   sccnkbdunlock(sc, >grab_state[lev]);
 }
 }
 
@@ -1763,6 +1790,7 @@ sc_cnungrab(struct consdev *cp)
 sc = sc_console->sc;
 lev = atomic_load_acq_int(>grab_level) - 1;
 if (lev >= 0 && lev < 2) {
+   sccnkbdlock(sc, >grab_state[lev]);
sccnscrlock(sc, >grab_state[lev]);
sccnclose(sc, >grab_state[lev]);
 }
@@ -1825,16 +1853,20 @@ sc_cnputc(struct consdev *cd, int c)
 static int
 sc_cngetc(struct consdev *cd)
 {
+struct sc_cnstate st;
 int c, s;
 
 /* assert(sc_console != NULL) */
+sccnopen(sc_console->sc, , 1);
 s = spltty();  /* block sckbdevent and scrn_timer while we poll */
-if (sc_console->sc->kbd == NULL) {
+if (!st.kbd_opened) {
splx(s);
-   return -1;
+   sccnclose(sc_console->sc, );
+   return -1;  /* means no keyboard since we fudged the locking */
 }
-c = sc_cngetc_locked(NULL);
+c = sc_cngetc_locked();
 splx(s);
+sccnclose(sc_console->sc, );
 return c;
 }
 
@@ -1858,7 +1890,7 @@ sc_cngetc_locked(struct sc_cnstate *sp)
 if

svn commit: r305120 - head/sys/arm/allwinner

2016-08-31 Thread Jared McNeill 
Author: jmcneill
Date: Wed Aug 31 10:45:53 2016
New Revision: 305120
URL: https://svnweb.freebsd.org/changeset/base/305120

Log:
  Add support for Allwinner A64 USB PHY.
  
  Reviewed by:  manu

Modified:
  head/sys/arm/allwinner/aw_usbphy.c

Modified: head/sys/arm/allwinner/aw_usbphy.c
==
--- head/sys/arm/allwinner/aw_usbphy.c  Wed Aug 31 10:45:33 2016
(r305119)
+++ head/sys/arm/allwinner/aw_usbphy.c  Wed Aug 31 10:45:53 2016
(r305120)
@@ -40,6 +40,7 @@ __FBSDID("$FreeBSD$");
 #include 
 #include 
 #include 
+#include 
 
 #include 
 #include 
@@ -53,25 +54,89 @@ __FBSDID("$FreeBSD$");
 #include "phy_if.h"
 
 #defineUSBPHY_NPHYS4
+#defineUSBPHY_NRES USBPHY_NPHYS
+
+enum awusbphy_type {
+   AWUSBPHY_TYPE_A10 = 1,
+   AWUSBPHY_TYPE_A13,
+   AWUSBPHY_TYPE_A20,
+   AWUSBPHY_TYPE_A31,
+   AWUSBPHY_TYPE_A83T,
+   AWUSBPHY_TYPE_H3,
+   AWUSBPHY_TYPE_A64
+};
 
 static struct ofw_compat_data compat_data[] = {
-   { "allwinner,sun4i-a10-usb-phy",1 },
-   { "allwinner,sun5i-a13-usb-phy",1 },
-   { "allwinner,sun6i-a31-usb-phy",1 },
-   { "allwinner,sun7i-a20-usb-phy",1 },
-   { "allwinner,sun8i-a83t-usb-phy",   1 },
-   { "allwinner,sun8i-h3-usb-phy", 1 },
+   { "allwinner,sun4i-a10-usb-phy",AWUSBPHY_TYPE_A10 },
+   { "allwinner,sun5i-a13-usb-phy",AWUSBPHY_TYPE_A13 },
+   { "allwinner,sun6i-a31-usb-phy",AWUSBPHY_TYPE_A31 },
+   { "allwinner,sun7i-a20-usb-phy",AWUSBPHY_TYPE_A20 },
+   { "allwinner,sun8i-a83t-usb-phy",   AWUSBPHY_TYPE_A83T },
+   { "allwinner,sun8i-h3-usb-phy", AWUSBPHY_TYPE_H3 },
+   { "allwinner,sun50i-a64-usb-phy",   AWUSBPHY_TYPE_A64 },
{ NULL, 0 }
 };
 
 struct awusbphy_softc {
+   struct resource *   res[USBPHY_NRES];
regulator_t reg[USBPHY_NPHYS];
gpio_pin_t  id_det_pin;
int id_det_valid;
gpio_pin_t  vbus_det_pin;
int vbus_det_valid;
+   enum awusbphy_type  phy_type;
 };
 
+static struct resource_spec awusbphy_spec[] = {
+   { SYS_RES_MEMORY,   0,  RF_ACTIVE },
+   { SYS_RES_MEMORY,   1,  RF_ACTIVE },
+   { SYS_RES_MEMORY,   2,  RF_ACTIVE | RF_OPTIONAL },
+   { SYS_RES_MEMORY,   3,  RF_ACTIVE | RF_OPTIONAL },
+   { -1, 0 }
+};
+
+#defineRD4(sc, i, o)   bus_read_4((sc)->res[(i)], (o))
+#defineWR4(sc, i, o, v)bus_write_4((sc)->res[(i)], (o), (v))
+#defineCLR4(sc, i, o, m)   WR4(sc, i, o, RD4(sc, i, o) & ~(m))
+#defineSET4(sc, i, o, m)   WR4(sc, i, o, RD4(sc, i, o) | (m))
+
+#defineOTG_PHY_CFG 0x20
+#define OTG_PHY_ROUTE_OTG  (1 << 0)
+#definePMU_IRQ_ENABLE  0x00
+#define PMU_AHB_INCR8  (1 << 10)
+#define PMU_AHB_INCR4  (1 << 9)
+#define PMU_AHB_INCRX_ALIGN(1 << 8)
+#define PMU_ULPI_BYPASS(1 << 0)
+#definePMU_UNK_H3  0x10
+#define PMU_UNK_H3_CLR 0x2
+
+static void
+awusbphy_configure(device_t dev, int phyno)
+{
+   struct awusbphy_softc *sc;
+
+   sc = device_get_softc(dev);
+
+   if (sc->res[phyno] == NULL)
+   return;
+
+   if (sc->phy_type == AWUSBPHY_TYPE_A64)  {
+   CLR4(sc, phyno, PMU_UNK_H3, PMU_UNK_H3_CLR);
+
+   /* EHCI0 and OTG share a PHY */
+   if (phyno == 0)
+   SET4(sc, 0, OTG_PHY_CFG, OTG_PHY_ROUTE_OTG);
+   else if (phyno == 1)
+   CLR4(sc, 0, OTG_PHY_CFG, OTG_PHY_ROUTE_OTG);
+   }
+
+   if (phyno > 0) {
+   /* Enable passby */
+   SET4(sc, phyno, PMU_IRQ_ENABLE, PMU_ULPI_BYPASS |
+   PMU_AHB_INCR8 | PMU_AHB_INCR4 | PMU_AHB_INCRX_ALIGN);
+   }
+}
+
 static int
 awusbphy_init(device_t dev)
 {
@@ -86,6 +151,8 @@ awusbphy_init(device_t dev)
sc = device_get_softc(dev);
node = ofw_bus_get_node(dev);
 
+   sc->phy_type = ofw_bus_search_compatible(dev, compat_data)->ocd_data;
+
/* Enable clocks */
for (off = 0; clk_get_by_ofw_index(dev, 0, off, ) == 0; off++) {
error = clk_enable(clk);
@@ -123,6 +190,10 @@ awusbphy_init(device_t dev)
if (error == 0)
sc->vbus_det_valid = 1;
 
+   /* Allocate resources */
+   if (bus_alloc_resources(dev, awusbphy_spec, sc->res) != 0)
+   device_printf(dev, "couldn't allocate resources\n");
+
return (0);
 }
 
@@ -159,6 +230,9 @@ awusbphy_phy_enable(device_t dev, intptr
 
sc = device_get_softc(dev);
 
+   /* Configure PHY */
+   awusbphy_configure(dev, phy);
+
/* Regulators are optional. If

svn commit: r305119 - in head/sys/arm/ti: am335x cpsw

2016-08-31 Thread Bjoern A. Zeeb 
Author: bz
Date: Wed Aug 31 10:45:33 2016
New Revision: 305119
URL: https://svnweb.freebsd.org/changeset/base/305119

Log:
  After r305113, try to properly replace the magic numbers with
  proper #defines for this driver (not using the wrong header).

Modified:
  head/sys/arm/ti/am335x/am335x_scm.h
  head/sys/arm/ti/cpsw/if_cpsw.c
  head/sys/arm/ti/cpsw/if_cpswreg.h

Modified: head/sys/arm/ti/am335x/am335x_scm.h
==
--- head/sys/arm/ti/am335x/am335x_scm.h Wed Aug 31 09:50:02 2016
(r305118)
+++ head/sys/arm/ti/am335x/am335x_scm.h Wed Aug 31 10:45:33 2016
(r305119)
@@ -42,8 +42,6 @@
 #defineSCM_USB_STS00x624
 #defineSCM_USB_CTRL1   0x628
 #defineSCM_USB_STS10x62C
-#defineSCM_MAC_ID0_LO  0x630
-#defineSCM_MAC_ID0_HI  0x634
 #defineSCM_PWMSS_CTRL  0x664
 
 #endif /* __AM335X_SCM_H__ */

Modified: head/sys/arm/ti/cpsw/if_cpsw.c
==
--- head/sys/arm/ti/cpsw/if_cpsw.c  Wed Aug 31 09:50:02 2016
(r305118)
+++ head/sys/arm/ti/cpsw/if_cpsw.c  Wed Aug 31 10:45:33 2016
(r305119)
@@ -1019,14 +1019,14 @@ cpswp_attach(device_t dev)
IFQ_SET_READY(>if_snd);
 
/* Get high part of MAC address from control module (mac_id[0|1]_hi) */
-   ti_scm_reg_read_4(SCM_MAC_ID0_HI + sc->unit * 8, );
+   ti_scm_reg_read_4(CPSW_MAC_ID0_HI + sc->unit * 8, );
mac_addr[0] = reg & 0xFF;
mac_addr[1] = (reg >>  8) & 0xFF;
mac_addr[2] = (reg >> 16) & 0xFF;
mac_addr[3] = (reg >> 24) & 0xFF;
 
/* Get low part of MAC address from control module (mac_id[0|1]_lo) */
-   ti_scm_reg_read_4(SCM_MAC_ID0_LO + sc->unit * 8, );
+   ti_scm_reg_read_4(CPSW_MAC_ID0_LO + sc->unit * 8, );
mac_addr[4] = reg & 0xFF;
mac_addr[5] = (reg >>  8) & 0xFF;
 

Modified: head/sys/arm/ti/cpsw/if_cpswreg.h
==
--- head/sys/arm/ti/cpsw/if_cpswreg.h   Wed Aug 31 09:50:02 2016
(r305118)
+++ head/sys/arm/ti/cpsw/if_cpswreg.h   Wed Aug 31 10:45:33 2016
(r305119)
@@ -46,6 +46,9 @@
 #defineCPSW_PORT_P_SA_LO(p)(CPSW_PORT_OFFSET + 0x120 + 
((p-1) * 0x100))
 #defineCPSW_PORT_P_SA_HI(p)(CPSW_PORT_OFFSET + 0x124 + 
((p-1) * 0x100))
 
+#defineCPSW_MAC_ID0_LO 0x0630
+#defineCPSW_MAC_ID0_HI 0x0634
+
 #defineCPSW_CPDMA_OFFSET   0x0800
 #defineCPSW_CPDMA_TX_CONTROL   (CPSW_CPDMA_OFFSET + 0x04)
 #defineCPSW_CPDMA_TX_TEARDOWN  (CPSW_CPDMA_OFFSET + 0x08)
___
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: r305113 - in head/sys/arm/ti: am335x cpsw

2016-08-31 Thread Bjoern A. Zeeb 

On 31 Aug 2016, at 7:03, Luiz Otavio O Souza wrote:


Author: loos
Date: Wed Aug 31 07:03:06 2016
New Revision: 305113
URL: https://svnweb.freebsd.org/changeset/base/305113

Log:
  Replace more magic numbers with the proper register names.

  Sponsored by: Rubicon Communications, LLC (Netgate)

Modified:
  head/sys/arm/ti/am335x/am335x_scm.h
  head/sys/arm/ti/cpsw/if_cpsw.c


Was this the right header file?  Looks like it wasn’t to me :(

/scratch/tmp/bz/head.svn/sys/modules/cpsw/../../arm/ti/cpsw/if_cpsw.c:1022:20: 
error: use of undeclared identifier 'SCM_MAC_ID0_HI'
/scratch/tmp/bz/head.svn/sys/modules/cpsw/../../arm/ti/cpsw/if_cpsw.c:1029:20: 
error: use of undeclared identifier 'SCM_MAC_ID0_LO'





Modified: head/sys/arm/ti/am335x/am335x_scm.h
==
--- head/sys/arm/ti/am335x/am335x_scm.h	Wed Aug 31 06:46:47 
2016	(r305112)
+++ head/sys/arm/ti/am335x/am335x_scm.h	Wed Aug 31 07:03:06 
2016	(r305113)

@@ -34,6 +34,8 @@
 #defineSCM_USB_STS00x624
 #defineSCM_USB_CTRL1   0x628
 #defineSCM_USB_STS10x62C
+#defineSCM_MAC_ID0_LO  0x630
+#defineSCM_MAC_ID0_HI  0x634
 #defineSCM_PWMSS_CTRL  0x664

 #endif /* __AM335X_SCM_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"

svn commit: r305118 - stable/10/sbin/camcontrol

2016-08-31 Thread Alexander Motin 
Author: mav
Date: Wed Aug 31 09:50:02 2016
New Revision: 305118
URL: https://svnweb.freebsd.org/changeset/base/305118

Log:
  MFC r304751:  Fix minor copy/paste bug.

Modified:
  stable/10/sbin/camcontrol/camcontrol.c
Directory Properties:
  stable/10/   (props changed)

Modified: stable/10/sbin/camcontrol/camcontrol.c
==
--- stable/10/sbin/camcontrol/camcontrol.c  Wed Aug 31 09:49:26 2016
(r305117)
+++ stable/10/sbin/camcontrol/camcontrol.c  Wed Aug 31 09:50:02 2016
(r305118)
@@ -7352,7 +7352,7 @@ smpphycontrol(struct cam_device *device,
response = malloc(sizeof(*response));
if (response == NULL) {
warn("%s: unable to allocate %zd bytes", __func__,
-sizeof(*request));
+sizeof(*response));
retval = 1;
goto bailout;
}
___
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: r305117 - stable/11/sbin/camcontrol

2016-08-31 Thread Alexander Motin 
Author: mav
Date: Wed Aug 31 09:49:26 2016
New Revision: 305117
URL: https://svnweb.freebsd.org/changeset/base/305117

Log:
  MFC r304751:  Fix minor copy/paste bug.

Modified:
  stable/11/sbin/camcontrol/camcontrol.c
Directory Properties:
  stable/11/   (props changed)

Modified: stable/11/sbin/camcontrol/camcontrol.c
==
--- stable/11/sbin/camcontrol/camcontrol.c  Wed Aug 31 09:23:09 2016
(r305116)
+++ stable/11/sbin/camcontrol/camcontrol.c  Wed Aug 31 09:49:26 2016
(r305117)
@@ -7515,7 +7515,7 @@ smpphycontrol(struct cam_device *device,
response = malloc(sizeof(*response));
if (response == NULL) {
warn("%s: unable to allocate %zd bytes", __func__,
-sizeof(*request));
+sizeof(*response));
retval = 1;
goto bailout;
}
___
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: r305116 - head/lib/libstand

2016-08-31 Thread Toomas Soome 
Author: tsoome
Date: Wed Aug 31 09:23:09 2016
New Revision: 305116
URL: https://svnweb.freebsd.org/changeset/base/305116

Log:
  recvtftp() is broken for large files, report file size
  
  The tftp download for large files will cause internal block id
  to wrap to 0 as the data type is unsigned short.
  
  Also provide file size information for stat.
  
  PR:   200500
  Reported by:  tsoome
  Reviewed by:  allanjude
  Approved by:  allanjude (mentor)
  Differential Revision:https://reviews.freebsd.org/D7660

Modified:
  head/lib/libstand/tftp.c

Modified: head/lib/libstand/tftp.c
==
--- head/lib/libstand/tftp.cWed Aug 31 07:42:46 2016(r305115)
+++ head/lib/libstand/tftp.cWed Aug 31 09:23:09 2016(r305116)
@@ -200,7 +200,7 @@ recvtftp(struct tftp_handle *h, void *pk
case DATA: {
int got;
 
-   if (htons(t->th_block) != d->xid) {
+   if (htons(t->th_block) != (u_short) d->xid) {
/*
 * Expected block?
 */
@@ -560,7 +560,7 @@ tftp_stat(struct open_file *f, struct st
sb->st_nlink = 1;
sb->st_uid = 0;
sb->st_gid = 0;
-   sb->st_size = -1;
+   sb->st_size = (off_t) tftpfile->tftp_tsize;
return (0);
 }
 
@@ -731,6 +731,8 @@ tftp_parse_oack(struct tftp_handle *h, c
} else if (strcasecmp(tftp_options[i], "tsize") == 0) {
if (i + 1 < option_idx)
tsize = strtol(tftp_options[i + 1], (char **)NULL, 10);
+   if (tsize != 0)
+   h->tftp_tsize = tsize;
} else {
/* Do not allow any options we did not expect to be ACKed. */
printf("unexpected tftp option '%s'\n", tftp_options[i]);
___
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: r305115 - head/sys/arm/ti/am335x

2016-08-31 Thread Luiz Otavio O Souza 
Author: loos
Date: Wed Aug 31 07:42:46 2016
New Revision: 305115
URL: https://svnweb.freebsd.org/changeset/base/305115

Log:
  Add a driver for the AM335x bandgap sensor, an on-die temperature sensor
  as part of the AM335x control module extension.
  
  TI says that the bandgap sensor is not very accurate on AM335x, but in our
  tests it seems to be a good reference for the SoC temperature.
  
  TI details:
  
http://processors.wiki.ti.com/index.php/AM335x_Thermal_Considerations#Measuring_Case_Temperature
  
  Sponsored by: Rubicon Communications, LLC (Netgate)

Added:
  head/sys/arm/ti/am335x/am335x_scm.c   (contents, props changed)
Modified:
  head/sys/arm/ti/am335x/am335x_scm.h
  head/sys/arm/ti/am335x/files.am335x

Added: head/sys/arm/ti/am335x/am335x_scm.c
==
--- /dev/null   00:00:00 1970   (empty, because file is newly added)
+++ head/sys/arm/ti/am335x/am335x_scm.c Wed Aug 31 07:42:46 2016
(r305115)
@@ -0,0 +1,169 @@
+/*-
+ * Copyright (c) 2016 Rubicon Communications, LLC (Netgate)
+ * 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 
+#include 
+#include 
+#include 
+
+#include 
+
+#include 
+#include 
+#include 
+
+#defineTZ_ZEROC2731
+
+struct am335x_scm_softc {
+   int sc_last_temp;
+   struct sysctl_oid   *sc_temp_oid;
+};
+
+static int
+am335x_scm_temp_sysctl(SYSCTL_HANDLER_ARGS)
+{
+   device_t dev;
+   int i, temp;
+   struct am335x_scm_softc *sc;
+   uint32_t reg;
+
+   dev = (device_t)arg1;
+   sc = device_get_softc(dev);
+
+   /* Read the temperature and convert to Kelvin. */
+   for(i = 50; i > 0; i--) {
+   ti_scm_reg_read_4(SCM_BGAP_CTRL, );
+   if ((reg & SCM_BGAP_EOCZ) == 0)
+   break;
+   DELAY(50);
+   }
+   if ((reg & SCM_BGAP_EOCZ) == 0) {
+   sc->sc_last_temp =
+   (reg >> SCM_BGAP_TEMP_SHIFT) & SCM_BGAP_TEMP_MASK;
+   sc->sc_last_temp *= 10;
+   }
+   temp = sc->sc_last_temp + TZ_ZEROC;
+
+   return (sysctl_handle_int(oidp, , 0, req));
+}
+
+static void
+am335x_scm_identify(driver_t *driver, device_t parent)
+{
+   device_t child;
+
+   /* AM335x only. */
+   if (ti_chip() != CHIP_AM335X)
+   return;
+
+   /* Make sure we attach only once. */
+   if (device_find_child(parent, "am335x_scm", -1) != NULL)
+   return;
+
+   child = device_add_child(parent, "am335x_scm", -1);
+   if (child == NULL)
+   device_printf(parent, "cannot add ti_scm child\n");
+}
+
+static int
+am335x_scm_probe(device_t dev)
+{
+
+   device_set_desc(dev, "AM335x Control Module Extension");
+
+   return (BUS_PROBE_DEFAULT);
+}
+
+static int
+am335x_scm_attach(device_t dev)
+{
+   struct am335x_scm_softc *sc;
+   struct sysctl_ctx_list *ctx;
+   struct sysctl_oid_list *tree;
+   uint32_t reg;
+
+   /* Set ADC to continous mode, clear output reset. */
+   reg = SCM_BGAP_CLRZ | SCM_BGAP_CONTCONV;
+   ti_scm_reg_write_4(SCM_BGAP_CTRL, reg);
+   /* Flush write. */
+   ti_scm_reg_read_4(SCM_BGAP_CTRL, );
+   /* Start the ADC conversion. */
+   reg = SCM_BGAP_CLRZ | SCM_BGAP_CONTCONV | SCM_BGAP_SOC;
+   ti_scm_reg_write_4(SCM_BGAP_CTRL, reg);
+
+   /* Temperature sysctl. */
+   sc = device_get_softc(dev);
+ctx = device_get_sysctl_ctx(dev);
+   tree = SYSCTL_CHILDREN(device_get_sysctl_tree(dev));
+   sc->sc_temp_oid = SYSCTL_ADD_PROC(ctx, tree, OID_AUTO,
+

svn commit: r305114 - head/sys/arm/ti

2016-08-31 Thread Luiz Otavio O Souza 
Author: loos
Date: Wed Aug 31 07:22:14 2016
New Revision: 305114
URL: https://svnweb.freebsd.org/changeset/base/305114

Log:
  Allow the use of control module extensions to cope with specific platform
  features.
  
  Sponsored by: Rubicon Communications, LLC (Netgate)

Modified:
  head/sys/arm/ti/ti_scm.c

Modified: head/sys/arm/ti/ti_scm.c
==
--- head/sys/arm/ti/ti_scm.cWed Aug 31 07:03:06 2016(r305113)
+++ head/sys/arm/ti/ti_scm.cWed Aug 31 07:22:14 2016(r305114)
@@ -131,7 +131,11 @@ ti_scm_attach(device_t dev)
 
ti_scm_sc = sc;
 
-   return (0);
+   /* Attach platform extensions, if any. */
+   bus_generic_probe(dev);
+   bus_enumerate_hinted_children(dev);
+
+   return (bus_generic_attach(dev));
 }
 
 int
___
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: r305084 - head/tools/regression/capsicum/syscalls

2016-08-31 Thread Mariusz Zaborski 
On Tue, Aug 30, 2016 at 06:04:50PM -0700, Ngie Cooper wrote:
> On Tue, Aug 30, 2016 at 12:58 PM, Mariusz Zaborski  
> wrote:
> > Author: oshogbo
> > Date: Tue Aug 30 19:58:41 2016
> > New Revision: 305084
> > URL: https://svnweb.freebsd.org/changeset/base/305084
> >
> > Log:
> >   Fix Capsicum syscalls test suite.
> 
> Do you plan on MFCing the change?
emaste@ suggested to change it to the kyua tests, what would you think about
MFCing after that?

> Thanks,
> -Ngie

Thanks,
-- 
Mariusz Zaborski
oshogbo//vx | http://oshogbo.vexillium.org
FreeBSD commiter| https://freebsd.org
Software developer  | http://wheelsystems.com
If it's not broken, let's fix it till it is!!1


signature.asc
Description: PGP signature

svn commit: r305113 - in head/sys/arm/ti: am335x cpsw

2016-08-31 Thread Luiz Otavio O Souza 
Author: loos
Date: Wed Aug 31 07:03:06 2016
New Revision: 305113
URL: https://svnweb.freebsd.org/changeset/base/305113

Log:
  Replace more magic numbers with the proper register names.
  
  Sponsored by: Rubicon Communications, LLC (Netgate)

Modified:
  head/sys/arm/ti/am335x/am335x_scm.h
  head/sys/arm/ti/cpsw/if_cpsw.c

Modified: head/sys/arm/ti/am335x/am335x_scm.h
==
--- head/sys/arm/ti/am335x/am335x_scm.h Wed Aug 31 06:46:47 2016
(r305112)
+++ head/sys/arm/ti/am335x/am335x_scm.h Wed Aug 31 07:03:06 2016
(r305113)
@@ -34,6 +34,8 @@
 #defineSCM_USB_STS00x624
 #defineSCM_USB_CTRL1   0x628
 #defineSCM_USB_STS10x62C
+#defineSCM_MAC_ID0_LO  0x630
+#defineSCM_MAC_ID0_HI  0x634
 #defineSCM_PWMSS_CTRL  0x664
 
 #endif /* __AM335X_SCM_H__ */

Modified: head/sys/arm/ti/cpsw/if_cpsw.c
==
--- head/sys/arm/ti/cpsw/if_cpsw.c  Wed Aug 31 06:46:47 2016
(r305112)
+++ head/sys/arm/ti/cpsw/if_cpsw.c  Wed Aug 31 07:03:06 2016
(r305113)
@@ -1019,14 +1019,14 @@ cpswp_attach(device_t dev)
IFQ_SET_READY(>if_snd);
 
/* Get high part of MAC address from control module (mac_id[0|1]_hi) */
-   ti_scm_reg_read_4(0x634 + sc->unit * 8, );
+   ti_scm_reg_read_4(SCM_MAC_ID0_HI + sc->unit * 8, );
mac_addr[0] = reg & 0xFF;
mac_addr[1] = (reg >>  8) & 0xFF;
mac_addr[2] = (reg >> 16) & 0xFF;
mac_addr[3] = (reg >> 24) & 0xFF;
 
/* Get low part of MAC address from control module (mac_id[0|1]_lo) */
-   ti_scm_reg_read_4(0x630 + sc->unit * 8, );
+   ti_scm_reg_read_4(SCM_MAC_ID0_LO + sc->unit * 8, );
mac_addr[4] = reg & 0xFF;
mac_addr[5] = (reg >>  8) & 0xFF;
 
___
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: r305112 - head/sys/arm/ti/am335x

2016-08-31 Thread Luiz Otavio O Souza 
Author: loos
Date: Wed Aug 31 06:46:47 2016
New Revision: 305112
URL: https://svnweb.freebsd.org/changeset/base/305112

Log:
  Replace a magic number with the proper register name.
  
  Sponsored by: Rubicon Communications, LLC (Netgate)

Modified:
  head/sys/arm/ti/am335x/am335x_prcm.c
  head/sys/arm/ti/am335x/am335x_scm.h

Modified: head/sys/arm/ti/am335x/am335x_prcm.c
==
--- head/sys/arm/ti/am335x/am335x_prcm.cWed Aug 31 06:00:20 2016
(r305111)
+++ head/sys/arm/ti/am335x/am335x_prcm.cWed Aug 31 06:46:47 2016
(r305112)
@@ -619,10 +619,9 @@ am335x_clk_get_sysclk_freq(struct ti_clo
 {
uint32_t ctrl_status;
 
-   /* Read the input clock freq from the control module */
-   /* control_status reg (0x40) */
-   if (ti_scm_reg_read_4(0x40, _status))
-   return ENXIO;
+   /* Read the input clock freq from the control module. */
+   if (ti_scm_reg_read_4(SCM_CTRL_STATUS, _status))
+   return (ENXIO);
 
switch ((ctrl_status>>22) & 0x3) {
case 0x0:

Modified: head/sys/arm/ti/am335x/am335x_scm.h
==
--- head/sys/arm/ti/am335x/am335x_scm.h Wed Aug 31 06:00:20 2016
(r305111)
+++ head/sys/arm/ti/am335x/am335x_scm.h Wed Aug 31 06:46:47 2016
(r305112)
@@ -29,6 +29,7 @@
 #define __AM335X_SCM_H__
 
 /* AM335x-specific registers for control module (scm) */
+#defineSCM_CTRL_STATUS 0x40
 #defineSCM_USB_CTRL0   0x620
 #defineSCM_USB_STS00x624
 #defineSCM_USB_CTRL1   0x628
___
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: r305091 - head/sys/sys

2016-08-31 Thread Bruce Evans 

On Wed, 31 Aug 2016, Mateusz Guzik wrote:


On Tue, Aug 30, 2016 at 03:35:39PM -0700, Conrad Meyer wrote:

On Tue, Aug 30, 2016 at 2:48 PM, Mateusz Guzik  wrote:

Author: mjg
Date: Tue Aug 30 21:48:10 2016
New Revision: 305091
URL: https://svnweb.freebsd.org/changeset/base/305091

Log:
  fd: simplify fd testing in fget_locked by casting to u_int

Modified:
  head/sys/sys/filedesc.h

Modified: head/sys/sys/filedesc.h
==
--- head/sys/sys/filedesc.h Tue Aug 30 21:43:57 2016(r305090)
+++ head/sys/sys/filedesc.h Tue Aug 30 21:48:10 2016(r305091)
@@ -201,7 +201,7 @@ fget_locked(struct filedesc *fdp, int fd

FILEDESC_LOCK_ASSERT(fdp);

-   if (fd < 0 || fd > fdp->fd_lastfile)
+   if ((u_int)fd > fdp->fd_lastfile)
return (NULL);

return (fdp->fd_ofiles[fd].fde_file);


This was correct.  Now it is obfuscated by a bogus cast.  Let the compiler
do a strength reduction to a single comparison operator iff that is best.
The cast is like using "register".  In 1980's code I wrote lots of such
unsigned casts since compilers were not so good and didn't do the strength
reduction.


I notice that fd_lastfile is an 'int'.  Won't this trigger warnings
about the differing signedness of the two sides of the comparison?
Should fd_lastfile just be u_int as well?  (If not (there is some
valid negative value), this change may be invalid.)


Ugh.  fd_lastfile has the correct for a file descriptor (int).  Making it
unsigned would just require many more bogus casts to break warnings when
comparing it with file descriptors.

Almost all uses of unsigned types are errors.  At best, the unsigned type
doubles the range of non-negative values and gives arithmetic modulo 2**N
instead of normal arithmetic.


It is -1 just after inception and is supposed to grow immediately after.
That is, the table with -1 should never be accessible.

But now that you mention it I agree this is bad style. This
unnecessarily differs from the check in fget_unlocked so I'll just unify
it.


Don't break the latter too.  I already had to fix it once there.  It had
2 commits just to move the bogus cast back and forth before I removed it.
following churning:
- created in r89969.  This used 1 bogus cast.  It was copied from a bad
  example in kern_descrip.c.  It did a redundant check that (fd < 0).  I
  had committed fixes for some bad examples like this in kern, but not
  this one.
- I complained about this to Albert, but he didn't understand and changed
  it to the above in r89969.  This works, but it is obscure.  It depends
  on both fd and fdp->fd_lastfile having type int.  So both operands are
  promoted to u_int and unsigned magic works.
- some compilers apparently object to the implicit conversion of the second
  operand.  So it was changed to use 2 bogus casts in r100072.  This works,
  but it still depends on both fd and fdp->fd_lastfile having type int.
  Otherwise the cast might truncate 1 or both of them, or if they are
  smaller than int (perhaps int32_t) then it isn't clear if the cast
  works as intended.
- I got tired of giving lessons and committed what I wanted in r126592.
  See the log message for r126592 for less details.

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: r305111 - head/sys/dev/hyperv/utilities

2016-08-31 Thread Sepherosa Ziehau 
Author: sephe
Date: Wed Aug 31 06:00:20 2016
New Revision: 305111
URL: https://svnweb.freebsd.org/changeset/base/305111

Log:
  hyperv/timesync: Rework time adjustment policy
  
  - By default, adjust time upon SYNC request.  It can be disabled
through hw.hvtimesync.ignore_sync_req.  SYNC request will be
sent by hypervisor the host is resumed, rebooted, etc.
  - By default, adjust time upon SAMPLE request, if there is 100ms
difference between VM time and hypervisor time.  This can be
disabled through hw.hvtimesync.sample_drift.
  
  And nuke the unnecessary task, since channel callback is running
  in a Hyper-V taskqueue nowadays.
  
  Submitted by: YanZhe Chen 
  Discussed with:   Dexuan Cui , Hongjiang Zhang 
, sephe
  MFC after:1 week
  Sponsored by: Microsoft
  Differential Revision:https://reviews.freebsd.org/D7707

Modified:
  head/sys/dev/hyperv/utilities/hv_timesync.c

Modified: head/sys/dev/hyperv/utilities/hv_timesync.c
==
--- head/sys/dev/hyperv/utilities/hv_timesync.c Wed Aug 31 05:27:30 2016
(r305110)
+++ head/sys/dev/hyperv/utilities/hv_timesync.c Wed Aug 31 06:00:20 2016
(r305111)
@@ -39,7 +39,7 @@
 #include 
 #include 
 #include 
-#include 
+#include 
 
 #include 
 #include 
@@ -52,11 +52,7 @@
 #define HV_ICTIMESYNCFLAG_SYNC  1
 #define HV_ICTIMESYNCFLAG_SAMPLE2
 #define HV_NANO_SEC_PER_SEC 10
-
-/* Time Sync data */
-typedef struct {
-   uint64_t data;
-} time_sync_data;
+#define HV_NANO_SEC_PER_MILLI_SEC   100
 
 static const struct vmbus_ic_desc vmbus_timesync_descs[] = {
{
@@ -75,22 +71,43 @@ struct hv_ictimesync_data {
uint8_t flags;
 } __packed;
 
-typedef struct hv_timesync_sc {
-   hv_util_sc  util_sc;
-   struct task task;
-   time_sync_data  time_msg;
-} hv_timesync_sc;
+/*
+ * Globals
+ */
+SYSCTL_NODE(_hw, OID_AUTO, hvtimesync, CTLFLAG_RW | CTLFLAG_MPSAFE, NULL,
+"Hyper-V timesync interface");
+
+/* Ignore the sync request when set to 1. */
+static int ignore_sync_req = 0;
+SYSCTL_INT(_hw_hvtimesync, OID_AUTO, ignore_sync_req, CTLFLAG_RWTUN,
+_sync_req, 0,
+"Ignore the sync request when set to 1.");
+
+/*
+ * Trigger sample sync when drift exceeds threshold (ms).
+ * Ignore the sample request when set to 0.
+ */
+static int sample_drift = 100;
+SYSCTL_INT(_hw_hvtimesync, OID_AUTO, sample_drift, CTLFLAG_RWTUN,
+_drift, 0,
+"Threshold that makes sample request trigger the sync.");
 
 /**
- * Set host time based on time sync message from host
+ * @brief Synchronize time with host after reboot, restore, etc.
+ *
+ * ICTIMESYNCFLAG_SYNC flag bit indicates reboot, restore events of the VM.
+ * After reboot the flag ICTIMESYNCFLAG_SYNC is included in the first time
+ * message after the timesync channel is opened. Since the hv_utils module is
+ * loaded after hv_vmbus, the first message is usually missed. The other
+ * thing is, systime is automatically set to emulated hardware clock which may
+ * not be UTC time or in the same time zone. So, to override these effects, we
+ * use the first 50 time samples for initial system time setting.
  */
-static void
-hv_set_host_time(void *context, int pending)
+static inline
+void hv_adj_guesttime(hv_util_sc *sc, uint64_t hosttime, uint8_t flags)
 {
-   hv_timesync_sc *softc = (hv_timesync_sc*)context;
-   uint64_t hosttime = softc->time_msg.data;
struct timespec guest_ts, host_ts;
-   uint64_t host_tns;
+   uint64_t host_tns, guest_tns;
int64_t diff;
int error;
 
@@ -99,37 +116,35 @@ hv_set_host_time(void *context, int pend
host_ts.tv_nsec = (long)(host_tns%HV_NANO_SEC_PER_SEC);
 
nanotime(_ts);
+   guest_tns = guest_ts.tv_sec * HV_NANO_SEC_PER_SEC + guest_ts.tv_nsec;
 
-   diff = (int64_t)host_ts.tv_sec - (int64_t)guest_ts.tv_sec;
+   if ((flags & HV_ICTIMESYNCFLAG_SYNC) != 0 && ignore_sync_req == 0) {
+   if (bootverbose) {
+   device_printf(sc->ic_dev, "handle sync request "
+   "{host: %ju, guest: %ju}\n",
+   (uintmax_t)host_tns, (uintmax_t)guest_tns);
+   }
 
-   /*
-* If host differs by 5 seconds then make the guest catch up
-*/
-   if (diff > 5 || diff < -5) {
error = kern_clock_settime(curthread, CLOCK_REALTIME,
_ts);
+   return;
}
-}
 
-/**
- * @brief Synchronize time with host after reboot, restore, etc.
- *
- * ICTIMESYNCFLAG_SYNC flag bit indicates reboot, restore events of the VM.
- * After reboot the flag ICTIMESYNCFLAG_SYNC is included in the first time
- * message after the timesync channel is opened. Since the hv_utils module is
- * loaded after hv_vmbus, the first message is usually missed. The other
- * thing is, systime is automatically set to emulated hardware clock which