jchandra [Fri, 4 Jun 2010 06:35:36 +0000 (06:35 +0000)]
Make vm_contig_grow_cache() extern, and use it when vm_phys_alloc_contig()
fails to allocate MIPS page table pages. The current usage of VM_WAIT in
case of vm_phys_alloc_contig() failure is not correct, because:
"There is no guarantee that any of the available free (or cached) pages
after the VM_WAIT will fall within the range of suitable physical
addresses. Every time this function sleeps and a single page is freed
(or cached) by someone else, this function will be reawakened. With
a little bad luck, you could spin indefinitely."
We also add low and high parameters to vm_contig_grow_cache() and
vm_contig_launder() so that we restrict vm_contig_launder() to the range
of pages we are interested in.
trasz [Thu, 3 Jun 2010 14:29:17 +0000 (14:29 +0000)]
_posix1e_acl_sort() never returns anything other than 0; change its
return type to void and update callers. This simplifies code and
fixes one place where the returned value was not actually checked.
marius [Thu, 3 Jun 2010 13:04:56 +0000 (13:04 +0000)]
- Mention that VTOC8 labels are found in Fujitsu SPARC64 machines as well.
- Add information regarding VTOC8 bootrstrap code and how it's handled with
r208777 in place.
- Document the mapping of partition types to VTOC8 tags.
- Add examples for VTOC8 to the respective section.
- Eliminated hard sentence breaks.
Reviewed by: marcel (slightly buggy version)
MFC after: 3 days
marius [Thu, 3 Jun 2010 13:01:55 +0000 (13:01 +0000)]
- In gpart_bootfile_read() fix an off-by-one error preventing the bootstrap
file to be of maximum size.
- Add special handling required for SMI/VTOC8 disklabel partcode, i.e. avoid
overwriting the label when writing the bootstrap code to the partition
starting at 0 and install it to all partitions when the -i option is omitted
just like geom_sunlabel(4) and sunlabel(8) do by default.
- Add missing prototypes.
- Add const where applicable.
kib [Thu, 3 Jun 2010 10:24:53 +0000 (10:24 +0000)]
Extend the scope of the lock on the quota file vnode in quotaon() to
cover the initial read by dqopen(). Assert that vnode is locked in
dqopen(). Remove VFS_LOCK_GIANT() from dqopen(), since quotaon() keeps
Giant locked if needed around the call.
kib [Thu, 3 Jun 2010 10:20:08 +0000 (10:20 +0000)]
Sometimes vnodes share the lock despite being different vnodes on
different mount points, e.g. the nullfs vnode and the covered vnode
from the lower filesystem. In this case, existing assertion in
vop_rename_pre() may be triggered.
Check for vnode locks equiality instead of the vnodes itself to
not trip over the situation.
alc [Thu, 3 Jun 2010 03:55:22 +0000 (03:55 +0000)]
In the unlikely event that pmap_ts_referenced() demoted five superpage
mappings to the same underlying physical page, the calling thread would be
left forever pinned to the same processor.
jmallett [Wed, 2 Jun 2010 21:15:00 +0000 (21:15 +0000)]
Fix build for O32 systems without a TARGET_CPUTYPE defined. We must default to
MIPS-III because FreeBSD relies on a number of MIPS-III features; the ABI
default would be MIPS-I which we don't intend to support. Our old default
before I switched to using the ABI default was MIPS32.
raj [Wed, 2 Jun 2010 17:17:45 +0000 (17:17 +0000)]
Import the common Flattened Device Tree infrastructure.
o fdtbus(4) - the main abstract bus driver for all FDT-compliant systems. This
is a direct replacement for the many incompatible bus drivers grouping
integrated peripherals on embedded platforms (like obio(4), ocpbus(4) etc.)
o simplebus(4) - bus driver representing ePAPR style 'simple-bus' node, which
is an umbrella device for most of the integrated peripherals on a typical
system-on-chip device.
o Other components (common routines library, PCI node processing helper
functions)
Reviewed by: imp
Sponsored by: The FreeBSD Foundation
alc [Wed, 2 Jun 2010 15:46:37 +0000 (15:46 +0000)]
Minimize the use of the page queues lock for synchronizing access to the
page's dirty field. With the exception of one case, access to this field
is now synchronized by the object lock.
jhb [Wed, 2 Jun 2010 15:09:36 +0000 (15:09 +0000)]
MFamd64: Add a new macro PCPU_XEN_FIELDS to hold XEN-specific per-CPU
fields that is always included in PCPU_MD_FIELDS. The macro is empty for
non-XEN kernels. This avoids duplicating non-XEN per-CPU fields in two
places. While here, remove several unused fields from the XEN-specific
structure.
jmallett [Wed, 2 Jun 2010 11:06:03 +0000 (11:06 +0000)]
Add/improve mips64r2, Octeon, n32 and n64 support in the toolchain.
o) Add TARGET_ABI to the MIPS toolchain build process. This sets the default
ABI to one of o32, n32 or n64. If it is not set, o32 is assumed as that is
the current default.
o) Set the default GCC cpu type to any specified TARGET_CPUTYPE. This is
necessary to have a working "cc" if e.g. mips64 is specified, as binutils
will refuse to link objects using different ISAs in some cases.
o) Add support for n32 and n64 ABIs to binutils and GCC.
o) Add additional required libgcc2 stubs for n32 and n64.
o) Add support for the "mips64r2" architecture to GCC. Add the "octeon"
o) When static linking, wrap default libraries in --start-group and
--end-group. This is required for static linking to work on n64 with the
interdependencies between libraries there. This is what other OSes that
support n64 seem to do, as well.
o) Fix our GCC spec to define __mips64 for 64-bit targets, not __mips64__, the
former being what libgcc, etc., check and the latter seemingly being a
misspelling of a hand merge from a Linux spec.
o) When no TARGET_CPUTYPE is specified at build time, make GCC take the default
ISA from the ABI. Our old defaults were too liberal and assumed that 64-bit
ABIs should default to the MIPS64 ISA and that 32-bit ABIs should default to
the MIPS32 ISA, when we are supporting or will support some systems based on
earlier 32-bit and 64-bit ISAs, most notably MIPS-III.
o) Merge a new opcode file (and support code) from a later version of binutils
and add flags and code necessary to support Octeon-specific instructions.
This should also make merging opcodes for other modern architectures easier.
kib [Wed, 2 Jun 2010 09:59:05 +0000 (09:59 +0000)]
Add a facility to dynamically adjust or unconfigure p1003_1b mib.
Use it to allow to tune sem_nsem_max at runtime, only when sem.ko
module is present in kernel.
Requested and tested by: amdmi3
Reviewed by: jhb
MFC after: 3 days
alc [Tue, 1 Jun 2010 19:56:02 +0000 (19:56 +0000)]
In the case that mmu_booke_enter_locked() is changing the attributes of a
mapping but not changing the physical page being mapped, the wrong flags
were being inspected in order to determine whether or not to flush the
instruction cache. The effect of looking at the wrong flags was that the
instruction cache was never being flushed.
rpaulo [Tue, 1 Jun 2010 15:33:10 +0000 (15:33 +0000)]
Bring in a couple of fixes from the Linux ath9k related to chip hangs.
While there, try to make the register write pattern look like what's
done by ath9k.
alc [Tue, 1 Jun 2010 05:18:48 +0000 (05:18 +0000)]
Merge portions of r208645 and supporting code from the i386 pmap:
When I pushed down the page queues lock into pmap_is_modified(), I created
an ordering dependence: A pmap operation that clears PG_WRITEABLE and calls
vm_page_dirty() must perform the call first. Otherwise, pmap_is_modified()
could return FALSE without acquiring the page queues lock because the page
is not (currently) writeable, and the caller to pmap_is_modified() might
believe that the page's dirty field is clear because it has not seen the
effect of the vm_page_dirty() call.
When I pushed down the page queues lock into pmap_is_modified(), I
overlooked one place where this ordering dependence is violated:
pmap_enter(). In a rare situation pmap_enter() can be called to replace a
dirty mapping to one page with a mapping to another page. (I say rare
because replacements generally occur as a result of a copy-on-write fault,
and so the old page is not dirty.) This change delays clearing PG_WRITEABLE
until after vm_page_dirty() has been called.
Fixing the ordering dependency also makes it easy to introduce a small
optimization: When pmap_enter() used to replace a mapping to one page with a
mapping to another page, it freed the pv entry for the first mapping and
later called the pv entry allocator for the new mapping. Now, pmap_enter()
attempts to recycle the old pv entry, saving two calls to the pv entry
allocator.
There is no point in setting PG_WRITEABLE on unmanaged pages, so don't.
pjd [Mon, 31 May 2010 23:17:45 +0000 (23:17 +0000)]
Fix a bug where resilver is not started automatically on pool import or load.
If disk was missing on pool load or import and on next pool load or import
it was present, resilver wasn't started automatically and ZFS reported all disks
as ONLINE and healthy. Then, when another disk died, pool became unaccessible,
because if it was 2-way mirror or RAIDZ1 two vdevs were out of sync.
To fix the problem, start resilver automatically on pool load or import.
alc [Mon, 31 May 2010 01:43:02 +0000 (01:43 +0000)]
Merge portions of r208645 and supporting code from the i386 pmap:
When I pushed down the page queues lock into pmap_is_modified(), I created
an ordering dependence: A pmap operation that clears PG_WRITEABLE and calls
vm_page_dirty() must perform the call first. Otherwise, pmap_is_modified()
could return FALSE without acquiring the page queues lock because the page
is not (currently) writeable, and the caller to pmap_is_modified() might
believe that the page's dirty field is clear because it has not seen the
effect of the vm_page_dirty() call.
When I pushed down the page queues lock into pmap_is_modified(), I
overlooked one place where this ordering dependence is violated:
pmap_enter(). In a rare situation pmap_enter() can be called to replace a
dirty mapping to one page with a mapping to another page. (I say rare
because replacements generally occur as a result of a copy-on-write fault,
and so the old page is not dirty.) This change delays clearing PG_WRITEABLE
until after vm_page_dirty() has been called.
Fixing the ordering dependency also makes it easy to introduce a small
optimization: When pmap_enter() used to replace a mapping to one page with a
mapping to another page, it freed the pv entry for the first mapping and
later called the pv entry allocator for the new mapping. Now, pmap_enter()
attempts to recycle the old pv entry, saving two calls to the pv entry
allocator.
alc [Sun, 30 May 2010 20:31:12 +0000 (20:31 +0000)]
Simplify the inner loop of get_pv_entry(): While iterating over the page's
pv list, there is no point in checking whether or not the pv list is empty,
wait instead until the loop completes.
alc [Sun, 30 May 2010 18:48:41 +0000 (18:48 +0000)]
Simplify the inner loop of pmap_collect(): While iterating over the page's
pv list, there is no point in checking whether or not the pv list is empty.
Instead, wait until the loop completes.
jilles [Sun, 30 May 2010 14:11:27 +0000 (14:11 +0000)]
sh: Change interaction of command substitution and here documents.
If a command substitution contains a newline token, this no longer starts
here documents of outer commands. This way, we follow POSIX's idea of the
command substitution being a separate script more closely. It also matches
other shells better and is consistent with newline characters in quotes not
starting here documents.
The extension tested in parser/heredoc3.0 ($(cat <<EOF)\ntext\nEOF\n)
continues to be supported.
In particular, this change allows things like
cat <<EOF && echo `pwd`
(a `` command substitution after a here document)
which formerly silently used an empty file as the here document, because the
EOF of the inner command "pwd" also forced an empty here document.
alc [Sun, 30 May 2010 04:44:32 +0000 (04:44 +0000)]
Merge various changes from i386/i386/pmap.c:
The remaining, unmerged portions of r175404
Retire PMAP_DIAGNOSTIC. Any useful diagnostics that were conditionally
compiled under PMAP_DIAGNOSTIC are now KASSERT()s. (Note: The kernel
option DIAGNOSTIC still disables inlining of certain pmap functions.)
Eliminate dead code from pmap_enter(). This code implemented an assertion.
On i386, an equivalent check is already implemented. However, on amd64,
a small change is required to implement an equivalent check.
Eliminate \n from a nearby panic string.
Use KASSERT() to reimplement pmap_copy()'s two assertions.
Merge portions of r177659
To date, we have assumed that the TLB will only set the PG_M bit in a
PTE if that PTE has the PG_RW bit set. However, this assumption does
not hold on recent processors from Intel. For example, consider a PTE
that has the PG_RW bit set but the PG_M bit clear. Suppose this PTE
is cached in the TLB and later the PG_RW bit is cleared in the PTE,
but the corresponding TLB entry is not (yet) invalidated.
Historically, upon a write access using this (stale) TLB entry, the
TLB would observe that the PG_RW bit had been cleared and initiate a
page fault, aborting the setting of the PG_M bit in the PTE. Now,
however, P4- and Core2-family processors will set the PG_M bit before
observing that the PG_RW bit is clear and initiating a page fault. In
other words, the write does not occur but the PG_M bit is still set.
The real impact of this difference is not that great. Specifically,
we should no longer assert that any PTE with the PG_M bit set must
also have the PG_RW bit set, and we should ignore the state of the
PG_M bit unless the PG_RW bit is set.
r208609
Defer freeing any page table pages in pmap_remove_all() until after the
page queues lock is released. This may reduce the amount of time that the
page queues lock is held by pmap_remove_all().
r208645
When I pushed down the page queues lock into pmap_is_modified(), I created
an ordering dependence: A pmap operation that clears PG_WRITEABLE and calls
vm_page_dirty() must perform the call first. Otherwise, pmap_is_modified()
could return FALSE without acquiring the page queues lock because the page
is not (currently) writeable, and the caller to pmap_is_modified() might
believe that the page's dirty field is clear because it has not seen the
effect of the vm_page_dirty() call.
When I pushed down the page queues lock into pmap_is_modified(), I
overlooked one place where this ordering dependence is violated:
pmap_enter(). In a rare situation pmap_enter() can be called to replace a
dirty mapping to one page with a mapping to another page. (I say rare
because replacements generally occur as a result of a copy-on-write fault,
and so the old page is not dirty.) This change delays clearing PG_WRITEABLE
until after vm_page_dirty() has been called.
Fixing the ordering dependency also makes it easy to introduce a small
optimization: When pmap_enter() used to replace a mapping to one page with a
mapping to another page, it freed the pv entry for the first mapping and
later called the pv entry allocator for the new mapping. Now, pmap_enter()
attempts to recycle the old pv entry, saving two calls to the pv entry
allocator.
There is no point in setting PG_WRITEABLE on unmanaged pages, so don't.
Update a comment to reflect this.
Tidy up the variable declarations at the start of pmap_enter().
jilles [Sat, 29 May 2010 20:24:01 +0000 (20:24 +0000)]
Clarify devfs manpages slightly.
mount(8): add xref to devfs(5)
devfs(5): change example to something more likely to be useful (it is not
necessary to mount a devfs on /dev manually, but for chroots/jails it is
often needed), mention since when devfs is preferred to device nodes on ufs
alc [Sat, 29 May 2010 17:10:45 +0000 (17:10 +0000)]
When I pushed down the page queues lock into pmap_is_modified(), I created
an ordering dependence: A pmap operation that clears PG_WRITEABLE and calls
vm_page_dirty() must perform the call first. Otherwise, pmap_is_modified()
could return FALSE without acquiring the page queues lock because the page
is not (currently) writeable, and the caller to pmap_is_modified() might
believe that the page's dirty field is clear because it has not seen the
effect of the vm_page_dirty() call.
When I pushed down the page queues lock into pmap_is_modified(), I
overlooked one place where this ordering dependence is violated:
pmap_enter(). In a rare situation pmap_enter() can be called to replace a
dirty mapping to one page with a mapping to another page. (I say rare
because replacements generally occur as a result of a copy-on-write fault,
and so the old page is not dirty.) This change delays clearing PG_WRITEABLE
until after vm_page_dirty() has been called.
Fixing the ordering dependency also makes it easy to introduce a small
optimization: When pmap_enter() used to replace a mapping to one page with a
mapping to another page, it freed the pv entry for the first mapping and
later called the pv entry allocator for the new mapping. Now, pmap_enter()
attempts to recycle the old pv entry, saving two calls to the pv entry
allocator.
There is no point in setting PG_WRITEABLE on unmanaged pages, so don't.
Update a comment to reflect this.
Tidy up the variable declarations at the start of pmap_enter().
rpaulo [Sat, 29 May 2010 16:14:02 +0000 (16:14 +0000)]
Due to the way HALDEBUG() is defined, we need to add curly brackets
when using it as a sole if clause instruction.
While there, fix 'const static' typo.
nyan [Sat, 29 May 2010 09:07:40 +0000 (09:07 +0000)]
MFi386: revision 178471
- Add an integer argument to idle to indicate how likely we are to wake
from idle over the next tick.
- Add a new MD routine, cpu_wake_idle() to wakeup idle threads who are
suspended in cpu specific states. This function can fail and cause the
scheduler to fall back to another mechanism (ipi).
- Implement support for mwait in cpu_idle() on i386/amd64 machines that
support it. mwait is a higher performance way to synchronize cpus
as compared to hlt & ipis.
- Allow selecting the idle routine by name via sysctl machdep.idle. This
replaces machdep.cpu_idle_hlt. Only idle routines supported by the
current machine are permitted.