svn commit: r262929 - head/sys/dev/ffec
Author: hrs
Date: Sat Mar 8 14:58:39 2014
New Revision: 262929
URL: http://svnweb.freebsd.org/changeset/base/262929
Log:
Fix another bug in multicast filtering. i.MX uses 6 bits from MSB in
LE CRC32 for the hash value, not the lowest 6 bits in BE CRC32.
Tested by:Takanori Sawada
PR: arm/187179
Modified:
head/sys/dev/ffec/if_ffec.c
Modified: head/sys/dev/ffec/if_ffec.c
==
--- head/sys/dev/ffec/if_ffec.c Sat Mar 8 12:26:17 2014(r262928)
+++ head/sys/dev/ffec/if_ffec.c Sat Mar 8 14:58:39 2014(r262929)
@@ -959,9 +959,10 @@ ffec_setup_rxfilter(struct ffec_softc *s
TAILQ_FOREACH(ifma, &sc->ifp->if_multiaddrs, ifma_link) {
if (ifma->ifma_addr->sa_family != AF_LINK)
continue;
- crc = ether_crc32_be(LLADDR((struct sockaddr_dl *)
+ /* 6 bits from MSB in LE CRC32 are used for hash. */
+ crc = ether_crc32_le(LLADDR((struct sockaddr_dl *)
ifma->ifma_addr), ETHER_ADDR_LEN);
- ghash |= 1LLU << (crc & 0x3f);
+ ghash |= 1LLU << (((uint8_t *)&crc)[3] >> 2);
}
if_maddr_runlock(ifp);
}
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
Re: svn commit: r262411 - head/sys/arm/arm
On Fri, 2014-03-07 at 17:21 +0200, Konstantin Belousov wrote: > On Wed, Mar 05, 2014 at 06:22:47AM -0700, Ian Lepore wrote: > > On Wed, 2014-03-05 at 13:54 +0200, Konstantin Belousov wrote: > > > On Sun, Feb 23, 2014 at 10:52:48PM +, Ian Lepore wrote: > > > > Author: ian > > > > Date: Sun Feb 23 22:52:48 2014 > > > > New Revision: 262411 > > > > URL: http://svnweb.freebsd.org/changeset/base/262411 > > > > > > > > Log: > > > > If the L2 cache type is PIPT, pass a physical address for a flush. > > > > > > > > While this is technically more correct, I don't think it much matters, > > > > because the only thing in the tree that calls cpu_flush_dcache() is > > > > md(4) > > > > and I'm > 99% sure it's bogus that it does so; md has no ability to do > > > > anything that can perturb data cache coherency. > > > > > > Yes, md(4) does not break data cache coherency, but I think that > > > Marcel added the flush to ensure instruction cache coherency. The > > > intent was to ensure that harward-architecture machines would > > > see up-to-date memory content when fetching instructions after > > > read on md(4). > > > > Oh. If that's necessary on ia64, it seems like ia64/elf_machdep.c would > > be the place to do the flush. > > I am not sure about ia64, it was needed for PowerPC, I think. > The issue is not limited to the module loads, so elf_machdep.c cannot > solve the problem. The part of the commit message for r192323 about ARM is just wrong. I don't know enough about the other platforms mentioned in there to comment on those, but the flush call which that commit message bills as avoiding pessimization just maximially pessimizes things on ARM. -- Ian ___ svn-src-head@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r262930 - head/sys/dev/ath
Author: rpaulo Date: Sat Mar 8 19:35:31 2014 New Revision: 262930 URL: http://svnweb.freebsd.org/changeset/base/262930 Log: Call ieee80211_dump_pkt() based on IFF_DUMPPKTS(). MFC after:3 days Modified: head/sys/dev/ath/if_ath_tx.c Modified: head/sys/dev/ath/if_ath_tx.c == --- head/sys/dev/ath/if_ath_tx.cSat Mar 8 14:58:39 2014 (r262929) +++ head/sys/dev/ath/if_ath_tx.cSat Mar 8 19:35:31 2014 (r262930) @@ -3769,9 +3769,10 @@ ath_tx_tid_drain_print(struct ath_softc ni->ni_txseqs[tid->tid]); /* XXX Dump the frame, see what it is? */ - ieee80211_dump_pkt(ni->ni_ic, - mtod(bf->bf_m, const uint8_t *), - bf->bf_m->m_len, 0, -1); + if (IFF_DUMPPKTS(sc, ATH_DEBUG_XMIT)) + ieee80211_dump_pkt(ni->ni_ic, + mtod(bf->bf_m, const uint8_t *), + bf->bf_m->m_len, 0, -1); } /* ___ svn-src-head@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r262931 - in head/bin: kill sh sh/tests/builtins
Author: jilles
Date: Sat Mar 8 19:44:34 2014
New Revision: 262931
URL: http://svnweb.freebsd.org/changeset/base/262931
Log:
sh: Successfully do nothing when killing a terminated job.
If a job has terminated but is still known, silently do nothing when using
the kill builtin with the job specifier. Formerly, the shell called kill()
with the process group ID that might have been reused.
Added:
head/bin/sh/tests/builtins/kill1.0 (contents, props changed)
Modified:
head/bin/kill/kill.c
head/bin/sh/jobs.c
head/bin/sh/tests/builtins/Makefile
Modified: head/bin/kill/kill.c
==
--- head/bin/kill/kill.cSat Mar 8 19:35:31 2014(r262930)
+++ head/bin/kill/kill.cSat Mar 8 19:44:34 2014(r262931)
@@ -133,9 +133,15 @@ main(int argc, char *argv[])
for (errors = 0; argc; argc--, argv++) {
#ifdef SHELL
- if (**argv == '%')
+ if (**argv == '%') {
pid = getjobpgrp(*argv);
- else
+ /*
+* Silently ignore terminated jobs, like the kernel
+* silently ignores zombies.
+*/
+ if (pid == 0)
+ continue;
+ } else
#endif
{
pid = strtol(*argv, &ep, 10);
Modified: head/bin/sh/jobs.c
==
--- head/bin/sh/jobs.c Sat Mar 8 19:35:31 2014(r262930)
+++ head/bin/sh/jobs.c Sat Mar 8 19:44:34 2014(r262931)
@@ -645,6 +645,8 @@ getjobpgrp(char *name)
struct job *jp;
jp = getjob(name);
+ if (jp->state == JOBDONE)
+ return 0;
return -jp->ps[0].pid;
}
Modified: head/bin/sh/tests/builtins/Makefile
==
--- head/bin/sh/tests/builtins/Makefile Sat Mar 8 19:35:31 2014
(r262930)
+++ head/bin/sh/tests/builtins/Makefile Sat Mar 8 19:44:34 2014
(r262931)
@@ -86,6 +86,7 @@ FILES+= hash3.0 hash3.0.stdout
FILES+=hash4.0
FILES+=jobid1.0
FILES+=jobid2.0
+FILES+=kill1.0
FILES+=lineno.0 lineno.0.stdout
FILES+=lineno2.0
FILES+=local1.0
Added: head/bin/sh/tests/builtins/kill1.0
==
--- /dev/null 00:00:00 1970 (empty, because file is newly added)
+++ head/bin/sh/tests/builtins/kill1.0 Sat Mar 8 19:44:34 2014
(r262931)
@@ -0,0 +1,8 @@
+# $FreeBSD$
+
+: &
+p1=$!
+: &
+p2=$!
+wait $p2
+kill %1
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r262932 - head/sys/dev/uart
Author: imp
Date: Sat Mar 8 19:59:52 2014
New Revision: 262932
URL: http://svnweb.freebsd.org/changeset/base/262932
Log:
Back out r262921. I don't know what I was thinking, but it is lame.
Modified:
head/sys/dev/uart/uart_core.c
Modified: head/sys/dev/uart/uart_core.c
==
--- head/sys/dev/uart/uart_core.c Sat Mar 8 19:44:34 2014
(r262931)
+++ head/sys/dev/uart/uart_core.c Sat Mar 8 19:59:52 2014
(r262932)
@@ -473,7 +473,7 @@ uart_bus_attach(device_t dev)
}
if (sc->sc_sysdev != NULL) {
- if (sc->sc_sysdev->baudrate != 0) {
+ if (sc->sc_sysdev->baudrate == 0) {
if (UART_IOCTL(sc, UART_IOCTL_BAUD,
(intptr_t)&sc->sc_sysdev->baudrate) != 0)
sc->sc_sysdev->baudrate = -1;
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r262934 - head/usr.bin/xinstall
Author: jilles
Date: Sat Mar 8 23:05:28 2014
New Revision: 262934
URL: http://svnweb.freebsd.org/changeset/base/262934
Log:
install: Use posix_spawnp() for starting strip and improve error messages.
Modified:
head/usr.bin/xinstall/xinstall.c
Modified: head/usr.bin/xinstall/xinstall.c
==
--- head/usr.bin/xinstall/xinstall.cSat Mar 8 20:31:04 2014
(r262933)
+++ head/usr.bin/xinstall/xinstall.cSat Mar 8 23:05:28 2014
(r262934)
@@ -62,6 +62,7 @@ __FBSDID("$FreeBSD$");
#include
#include
#include
+#include
#include
#include
#include
@@ -102,6 +103,8 @@ static enum {
DIGEST_SHA512,
} digesttype = DIGEST_NONE;
+extern char **environ;
+
static gid_t gid;
static uid_t uid;
static int dobackup, docompare, dodir, dolink, dopreserve, dostrip, dounpriv,
@@ -1215,27 +1218,33 @@ static void
strip(const char *to_name)
{
const char *stripbin;
- int serrno, status;
-
- switch (fork()) {
- case -1:
- serrno = errno;
+ const char *args[3];
+ pid_t pid;
+ int error, status;
+
+ stripbin = getenv("STRIPBIN");
+ if (stripbin == NULL)
+ stripbin = "strip";
+ args[0] = stripbin;
+ args[1] = to_name;
+ args[2] = NULL;
+ error = posix_spawnp(&pid, stripbin, NULL, NULL,
+ __DECONST(char **, args), environ);
+ if (error != 0) {
(void)unlink(to_name);
- errno = serrno;
- err(EX_TEMPFAIL, "fork");
- case 0:
- stripbin = getenv("STRIPBIN");
- if (stripbin == NULL)
- stripbin = "strip";
- execlp(stripbin, stripbin, to_name, (char *)NULL);
- err(EX_OSERR, "exec(%s)", stripbin);
- default:
- if (wait(&status) == -1 || status) {
- serrno = errno;
- (void)unlink(to_name);
- errc(EX_SOFTWARE, serrno, "wait");
- /* NOTREACHED */
- }
+ errc(error == EAGAIN || error == EPROCLIM || error == ENOMEM ?
+ EX_TEMPFAIL : EX_OSERR, error, "spawn %s", stripbin);
+ }
+ if (waitpid(pid, &status, 0) == -1) {
+ error = errno;
+ (void)unlink(to_name);
+ errc(EX_SOFTWARE, error, "wait");
+ /* NOTREACHED */
+ }
+ if (status != 0) {
+ (void)unlink(to_name);
+ errx(EX_SOFTWARE, "strip command %s failed on %s",
+ stripbin, to_name);
}
}
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r262935 - head/sys/arm/allwinner/a20
Author: ian Date: Sun Mar 9 01:52:21 2014 New Revision: 262935 URL: http://svnweb.freebsd.org/changeset/base/262935 Log: Follow r262916 with one more config file that references a renamed common.c Modified: head/sys/arm/allwinner/a20/files.a20 Modified: head/sys/arm/allwinner/a20/files.a20 == --- head/sys/arm/allwinner/a20/files.a20Sat Mar 8 23:05:28 2014 (r262934) +++ head/sys/arm/allwinner/a20/files.a20Sun Mar 9 01:52:21 2014 (r262935) @@ -19,5 +19,5 @@ arm/allwinner/if_emac.c optional emac arm/allwinner/a10_wdog.c standard arm/allwinner/timer.c standard arm/arm/bus_space-v6.c standard -arm/allwinner/common.c standard +arm/allwinner/a10_common.c standard arm/allwinner/a10_machdep.cstandard ___ svn-src-head@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r262936 - head/release/tools/arm
Author: gjb Date: Sun Mar 9 02:00:48 2014 New Revision: 262936 URL: http://svnweb.freebsd.org/changeset/base/262936 Log: Update board definition for WANDBOARD-QUAD to match crochet. Sponsored by: The FreeBSD Foundation Modified: head/release/tools/arm/crochet-WANDBOARD-QUAD.conf Modified: head/release/tools/arm/crochet-WANDBOARD-QUAD.conf == --- head/release/tools/arm/crochet-WANDBOARD-QUAD.conf Sun Mar 9 01:52:21 2014(r262935) +++ head/release/tools/arm/crochet-WANDBOARD-QUAD.conf Sun Mar 9 02:00:48 2014(r262936) @@ -5,7 +5,7 @@ # This is the configuration file for use with crochet to produce # FreeBSD WandboardQuad images. -board_setup WandboardQuad +board_setup Wandboard option ImageSize 1gb option AutoSize ___ svn-src-head@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r262937 - head/sys/dev/bvm
Author: neel
Date: Sun Mar 9 02:07:15 2014
New Revision: 262937
URL: http://svnweb.freebsd.org/changeset/base/262937
Log:
Fix panic "_mtx_lock_sleep: recursed on non-recursive mutex ttymtx" caused
when
acquiring the tty lock in bvm_tty_close(). Instead just assert that the tty
lock is held before calling callout_reset().
Modified:
head/sys/dev/bvm/bvm_console.c
Modified: head/sys/dev/bvm/bvm_console.c
==
--- head/sys/dev/bvm/bvm_console.c Sun Mar 9 02:00:48 2014
(r262936)
+++ head/sys/dev/bvm/bvm_console.c Sun Mar 9 02:07:15 2014
(r262937)
@@ -128,9 +128,8 @@ static void
bvm_tty_close(struct tty *tp)
{
- tty_lock(tp);
+ tty_lock_assert(tp, MA_OWNED);
callout_stop(&bvm_timer);
- tty_unlock(tp);
}
static void
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r262940 - head/sys/arm/at91
Author: imp Date: Sun Mar 9 02:28:30 2014 New Revision: 262940 URL: http://svnweb.freebsd.org/changeset/base/262940 Log: Remove bogus define that causes compile errors. Most of the defines for SAM9X are going away soonish anyway (once FDT works), but until then... Modified: head/sys/arm/at91/at91sam9x5reg.h Modified: head/sys/arm/at91/at91sam9x5reg.h == --- head/sys/arm/at91/at91sam9x5reg.h Sun Mar 9 02:11:53 2014 (r262939) +++ head/sys/arm/at91/at91sam9x5reg.h Sun Mar 9 02:28:30 2014 (r262940) @@ -221,7 +221,6 @@ #define AT91SAM9X25_IRQ_PIOCAT91SAM9X25_IRQ_PIOCD #define AT91SAM9X25_IRQ_PIODAT91SAM9X25_IRQ_PIOCD #define AT91SAM9X25_IRQ_NAND (-1) -#define AT91SAM9X25_IRQ_AIC(-1) #define AT91SAM9X25_AIC_BASE 0x000 #define AT91SAM9X25_AIC_SIZE 0x200 ___ svn-src-head@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r262941 - in head/sys/arm: arm include
Author: ian
Date: Sun Mar 9 03:00:03 2014
New Revision: 262941
URL: http://svnweb.freebsd.org/changeset/base/262941
Log:
Rework the VFP code that handles demand-based save and restore of state.
The old code was full of complexity that would only matter if the
kernel itself used the VFP hardware. Now that's reduced to either killing
the userland process or panicking the kernel on an illegal VFP instruction.
This removes most of the complexity from the assembler code, reducing it
to just calling the save code if the outgoing thread used the VFP.
The routine that stores the VFP state now takes a flag that indicates
whether the hardware should be disabled after saving state. Right now it
always is, but this makes the code ready to be used by get/set_mcontext()
(doing so will be addressed in a future commit).
Remove the arm-specific pc_vfpcthread from struct pcpu and use the MI
field pc_fpcurthread instead.
Reviewed by: cognet
Modified:
head/sys/arm/arm/genassym.c
head/sys/arm/arm/swtch.S
head/sys/arm/arm/vfp.c
head/sys/arm/include/pcpu.h
head/sys/arm/include/vfp.h
Modified: head/sys/arm/arm/genassym.c
==
--- head/sys/arm/arm/genassym.c Sun Mar 9 02:28:30 2014(r262940)
+++ head/sys/arm/arm/genassym.c Sun Mar 9 03:00:03 2014(r262941)
@@ -119,9 +119,7 @@ ASSYM(ARM_RAS_END, ARM_RAS_END);
#ifdef VFP
ASSYM(PCB_VFPSTATE, offsetof(struct pcb, pcb_vfpstate));
-ASSYM(PCB_VFPCPU, offsetof(struct pcb, pcb_vfpcpu));
-ASSYM(PC_VFPCTHREAD, offsetof(struct pcpu, pc_vfpcthread));
ASSYM(PC_CPU, offsetof(struct pcpu, pc_cpu));
ASSYM(PC_CURPMAP, offsetof(struct pcpu, pc_curpmap));
Modified: head/sys/arm/arm/swtch.S
==
--- head/sys/arm/arm/swtch.SSun Mar 9 02:28:30 2014(r262940)
+++ head/sys/arm/arm/swtch.SSun Mar 9 03:00:03 2014(r262941)
@@ -84,6 +84,8 @@
#include
#include
#include
+#include
+
__FBSDID("$FreeBSD$");
#define DOMAIN_CLIENT 0x01
@@ -102,6 +104,10 @@ __FBSDID("$FreeBSD$");
ldr tmp, .Lcurpcpu
#endif
+#ifdef VFP
+ .fpu vfp/* allow VFP instructions */
+#endif
+
.Lcurpcpu:
.word _C_LABEL(__pcpu)
.word PCPU_SIZE
@@ -121,16 +127,10 @@ ENTRY(cpu_throw)
GET_PCPU(r7, r9)
#ifdef VFP
- /*
-* vfp_discard will clear pcpu->pc_vfpcthread, and modify
-* and modify the control as needed.
-*/
- ldr r4, [r7, #(PC_VFPCTHREAD)] /* this thread using vfp? */
- cmp r0, r4
- bne 3f
- bl _C_LABEL(vfp_discard) /* yes, shut down vfp */
-3:
-#endif /* VFP */
+ fmrxr0, fpexc /* This thread is dying, if the VFP */
+ tst r0, #(VFPEXC_EN)/* is enabled, go shut it down */
+ blne_C_LABEL(vfp_discard) /* without preserving its state. */
+#endif
ldr r7, [r5, #(TD_PCB)] /* r7 = new thread's PCB */
@@ -319,30 +319,14 @@ ENTRY(cpu_switch)
/* rem: interrupts are enabled */
#ifdef VFP
- /*
-* vfp_store will clear pcpu->pc_vfpcthread, save
-* registers and state, and modify the control as needed.
-* a future exception will bounce the backup settings in the fp unit.
-* XXX vfp_store can't change r4
-*/
- GET_PCPU(r7, r8)
- ldr r8, [r7, #(PC_VFPCTHREAD)]
- cmp r4, r8 /* old thread used vfp? */
- bne 1f /* no, don't save */
- cmp r1, r4 /* same thread ? */
- beq 1f /* yes, skip vfp store */
-#ifdef SMP
- ldr r8, [r7, #(PC_CPU)] /* last used on this cpu? */
- ldr r3, [r2, #(PCB_VFPCPU)]
- cmp r8, r3 /* last cpu to use these registers? */
- bne 1f /* no. these values are stale */
+ fmrxr0, fpexc /* If the VFP is enabled */
+ tst r0, #(VFPEXC_EN)/* the current thread has */
+ movne r1, #1 /* used it, so go save */
+ addne r0, r2, #(PCB_VFPSTATE) /* the state into the PCB */
+ blne_C_LABEL(vfp_store) /* and disable the VFP. */
#endif
- add r0, r2, #(PCB_VFPSTATE)
- bl _C_LABEL(vfp_store)
-1:
-#endif /* VFP */
- /* r1 now free! */
+ /* r0-r3 now free! */
/* Third phase : restore saved context */
@@ -520,26 +504,12 @@ ENTRY(savectx)
add r2, r0, #(PCB_R8)
stmia r2, {r8-r13}
#ifdef VFP
- /*
-* vfp_store will clear pcpu->pc_vfpcthread, save
-* registers and state, and modify the control as needed.
-* a future exception will bounce the backup settings in the fp unit.
-*/
svn commit: r262942 - in head/sys/arm: arm include
Author: ian Date: Sun Mar 9 03:49:06 2014 New Revision: 262942 URL: http://svnweb.freebsd.org/changeset/base/262942 Log: Remove all dregs of a per-thread undefined-exception-mode stack. This is a leftover from the days when a low-level debugger had hooks in the undefined exception vector and needed stack space to function. These days it effectively isn't used because we switch immediately to the svc32 mode stack on exception entry. For that, the single undef mode stack per core that gets set up at init time works fine. The stack wasn't necessary but it was harmful, because the space for it was carved out of the normal per-thread svc32 stack, in effect cutting that 8K stack in half. If svc32 mode used more than 4k of stack space it wandered down into the undef mode stack, and then an undef exception would overwrite a couple words on the stack while switching to svc32 mode, corrupting the scv32 stack. Having another stack abut the bottom of the svc32 stack also effectively mooted the guard page below the stack. This work is based on analysis and patches submitted by Juergen Weiss. Modified: head/sys/arm/arm/genassym.c head/sys/arm/arm/machdep.c head/sys/arm/arm/swtch.S head/sys/arm/arm/vm_machdep.c head/sys/arm/include/param.h head/sys/arm/include/pcb.h Modified: head/sys/arm/arm/genassym.c == --- head/sys/arm/arm/genassym.c Sun Mar 9 03:00:03 2014(r262941) +++ head/sys/arm/arm/genassym.c Sun Mar 9 03:49:06 2014(r262942) @@ -60,7 +60,6 @@ ASSYM(PCB_NOALIGNFLT, PCB_NOALIGNFLT); ASSYM(PCB_ONFAULT, offsetof(struct pcb, pcb_onfault)); ASSYM(PCB_DACR, offsetof(struct pcb, pcb_dacr)); ASSYM(PCB_FLAGS, offsetof(struct pcb, pcb_flags)); -ASSYM(PCB_UND_SP, offsetof(struct pcb, un_32.pcb32_und_sp)); ASSYM(PCB_PAGEDIR, offsetof(struct pcb, pcb_pagedir)); ASSYM(PCB_L1VEC, offsetof(struct pcb, pcb_l1vec)); ASSYM(PCB_PL1VEC, offsetof(struct pcb, pcb_pl1vec)); Modified: head/sys/arm/arm/machdep.c == --- head/sys/arm/arm/machdep.c Sun Mar 9 03:00:03 2014(r262941) +++ head/sys/arm/arm/machdep.c Sun Mar 9 03:49:06 2014(r262942) @@ -379,8 +379,6 @@ cpu_startup(void *dummy) bufinit(); vm_pager_bufferinit(); - pcb->un_32.pcb32_und_sp = (u_int)thread0.td_kstack + - USPACE_UNDEF_STACK_TOP; pcb->un_32.pcb32_sp = (u_int)thread0.td_kstack + USPACE_SVC_STACK_TOP; vector_page_setprot(VM_PROT_READ); Modified: head/sys/arm/arm/swtch.S == --- head/sys/arm/arm/swtch.SSun Mar 9 03:00:03 2014(r262941) +++ head/sys/arm/arm/swtch.SSun Mar 9 03:49:06 2014(r262942) @@ -303,17 +303,6 @@ ENTRY(cpu_switch) /* Get the user structure for the new process in r9 */ ldr r9, [r1, #(TD_PCB)] -mrsr3, cpsr - /* -* We can do that, since -* PSR_SVC32_MODE|PSR_UND32_MODE == MSR_UND32_MODE -*/ - orr r8, r3, #(PSR_UND32_MODE) -msrcpsr_c, r8 - - str sp, [r2, #(PCB_UND_SP)] - -msrcpsr_c, r3 /* Restore the old mode */ /* rem: r2 = old PCB */ /* rem: r9 = new PCB */ /* rem: interrupts are enabled */ @@ -422,10 +411,6 @@ ENTRY(cpu_switch) movne r0, #0 /* We *know* vector_page's VA is 0x0 */ movne lr, pc ldrne pc, [r10, #CF_TLB_FLUSHID_SE] - /* -* We can do that, since -* PSR_SVC32_MODE|PSR_UND32_MODE == MSR_UND32_MODE -*/ .Lcs_context_switched: @@ -444,17 +429,6 @@ ENTRY(cpu_switch) /* rem: r9 = new PCB */ -mrsr3, cpsr - /* -* We can do that, since -* PSR_SVC32_MODE|PSR_UND32_MODE == MSR_UND32_MODE -*/ - orr r2, r3, #(PSR_UND32_MODE) - msr cpsr_c, r2 - - ldr sp, [r9, #(PCB_UND_SP)] - -msrcpsr_c, r3 /* Restore the old mode */ /* Restore all the save registers */ #ifndef _ARM_ARCH_5E add r7, r9, #PCB_R8 Modified: head/sys/arm/arm/vm_machdep.c == --- head/sys/arm/arm/vm_machdep.c Sun Mar 9 03:00:03 2014 (r262941) +++ head/sys/arm/arm/vm_machdep.c Sun Mar 9 03:49:06 2014 (r262942) @@ -144,7 +144,6 @@ cpu_fork(register struct thread *td1, re bcopy(td1->td_pcb, pcb2, sizeof(*pcb2)); mdp2 = &p2->p_md; bcopy(&td1->td_proc->p_md, mdp2, sizeof(*mdp2)); - pcb2->un_32.pcb32_und_sp = td2->td_kstack + USPACE_UNDEF_STACK_TOP; pcb2->un_32.pcb32_sp = td2->td_kstack + USPACE_SVC_STACK_TOP - sizeof(*pcb2); pmap_activate(td2); @@ -366,7 +365,6 @@ cpu_set_upcall(struc
svn commit: r262945 - head/lib/libcrypt
Author: jmg
Date: Sun Mar 9 07:45:59 2014
New Revision: 262945
URL: http://svnweb.freebsd.org/changeset/base/262945
Log:
various formating fixes, use NUL for NUL bytes..
drop out dated perf numbers (can't imagine people are still running
Pentium MMX 166's anymore)...
bump date...
drop max length of salt of 8 since _PASSWORD_LEN is now large, 128..
and state the max length of the salt depends upon the module,
sha-{256,512} have a max salt of 16..
recommend 8 characters of salt instead of just 2...
MFC after:1 week
Modified:
head/lib/libcrypt/crypt.3
Modified: head/lib/libcrypt/crypt.3
==
--- head/lib/libcrypt/crypt.3 Sun Mar 9 04:17:04 2014(r262944)
+++ head/lib/libcrypt/crypt.3 Sun Mar 9 07:45:59 2014(r262945)
@@ -29,7 +29,7 @@
.\"
.\" $FreeBSD$
.\"
-.Dd April 9, 2011
+.Dd March 9, 2014
.Dt CRYPT 3
.Os
.Sh NAME
@@ -63,11 +63,16 @@ Currently these include the
.Tn MD5
hash,
.Tn NT-Hash
-(compatible with Microsoft's NT scheme)
+.Pq compatible with Microsoft's NT scheme
and
.Tn Blowfish .
-The algorithm used will depend upon the format of the Salt (following
-the Modular Crypt Format (MCF)), if
+The algorithm used will depend upon the format of the Salt
+.Po
+following
+the Modular Crypt Format
+.Pq MCF
+.Pc ,
+if
.Tn DES
and/or
.Tn Blowfish
@@ -77,8 +82,10 @@ has been called to change the default.
.Pp
The first argument to
.Nm
-is the data to hash (usually a password), in a
-.Dv null Ns -terminated
+is the data to hash
+.Pq usually a password ,
+in a
+.Dv NUL Ns -terminated
string.
The second is the salt, in one of three forms:
.Pp
@@ -96,22 +103,19 @@ If it begins with the string
then the Modular Crypt Format is used, as outlined below.
.It Traditional
If neither of the above is true, it assumes the Traditional Format,
-using the entire string as the salt (or the first portion).
+using the entire string as the salt
+.Pq or the first portion .
.El
.Pp
All routines are designed to be time-consuming.
-A brief test on a
-.Tn Pentium
-166/MMX shows the
-.Tn DES
-crypt to do approximately 2640 crypts
-a CPU second and MD5 to do about 62 crypts a CPU second.
.Ss DES Extended Format:
The
.Ar key
-is divided into groups of 8 characters (the last group is null-padded)
-and the low-order 7 bits of each character (56 bits per group) are
-used to form the
+is divided into groups of 8 characters
+.Pq the last group is NUL-padded
+and the low-order 7 bits of each character
+.Pq 56 bits per group
+are used to form the
.Tn DES
key as follows:
the first group of 56 bits becomes the initial
@@ -127,7 +131,8 @@ The salt is a 9-character array consisti
by 4 bytes of iteration count and 4 bytes of salt.
These are encoded as printable characters, 6 bits per character,
least significant character first.
-The values 0 to 63 are encoded as ``./0-9A-Za-z''.
+The values 0 to 63 are encoded as
+.Dq ./0-9A-Za-z .
This allows 24 bits for both
.Fa count
and
@@ -138,7 +143,8 @@ The
introduces disorder in the
.Tn DES
algorithm in one of 16777216 or 4096 possible ways
-(i.e., with 24 or 12 bits: if bit
+.Po
+i.e., with 24 or 12 bits: if bit
.Em i
of the
.Ar salt
@@ -148,7 +154,8 @@ and
.Em i+24
are swapped in the
.Tn DES
-E-box output).
+E-box output
+.Pc .
.Pp
The
.Tn DES
@@ -157,11 +164,13 @@ key is used to encrypt a 64-bit constant
iterations of
.Tn DES .
The value returned is a
-.Dv null Ns -terminated
-string, 20 or 13 bytes (plus null) in length, consisting of the
+.Dv NUL Ns -terminated
+string, 20 or 13 bytes
+.Pq plus NUL
+in length, consisting of the
.Ar salt
followed by the encoded 64-bit encryption.
-.Ss "Modular" crypt:
+.Ss Modular crypt:
If the salt begins with the string
.Fa $digit$
then the Modular Crypt Format is used.
@@ -170,11 +179,10 @@ The
represents which algorithm is used in encryption.
Following the token is
the actual salt to use in the encryption.
-The length of the salt is limited
-to 8 characters--because the length of the returned output is also limited
-(_PASSWORD_LEN).
-The salt must be terminated with the end of the string
-(NULL) or a dollar sign.
+The maximum length of the salt used depends upon the module.
+The salt must be terminated with the end of the string character
+.Pq NUL
+or a dollar sign.
Any characters after the dollar sign are ignored.
.Pp
Currently supported algorithms are:
@@ -199,7 +207,7 @@ An example salt would be:
.Bl -tag -width 6n -offset indent
.It Cm "$4$thesalt$rest"
.El
-.Ss "Traditional" crypt:
+.Ss Traditional crypt:
The algorithm used will depend upon whether
.Fn crypt_set_format
has been called and whether a global default format has been specified.
@@ -216,7 +224,7 @@ if it is available, or MD5 if not.
.Pp
How the salt is used will depend upon the algorithm for the hash.
For
-best results, specify at least two characters of salt.
+best results, specify at least eight
