Roman Divacky [Wed, 18 Sep 2013 17:56:04 +0000 (17:56 +0000)]
Implement epoll support in Linuxulator. This is a tiny wrapper around kqueue
to implement epoll subset of functionality. The kqueue user data are 32bit
on i386 which is not enough for epoll user data so this patch overrides
kqueue fileops to maintain enough space in struct file.
Initial patch developed by me in 2007 and then extended and finished
by Yuri Victorovich.
Approved by: re (delphij)
Sponsored by: Google Summer of Code
Submitted by: Yuri Victorovich <yuri at rawbw dot com>
Tested by: Yuri Victorovich <yuri at rawbw dot com>
Fix 'make installcheck' to check for missing UID/GID as well, broken
since r249893, by adding a separate _installcheck_world and
_installcheck_kernel so the destination targets can be more explicit
on which they are needed for.
installcheck will call both, while installworld only calls
_installcheck_world and installkernel only calls _installcheck_kernel
While here, mark the internal targets as starting with _.
Reported by: des
Reviewed by: des
Pointyhat to: bdrewery
Approved by: re (delphij)
Disallow cap_enter() in freebsd32 compatibility mode.
The freebsd32 compatibility mode (for running 32-bit binaries on 64-bit
kernels) does not currently allow any system calls in capability mode, but
still permits cap_enter(). As a result, 32-bit binaries on 64-bit kernels
that use capability mode do not work (they crash after being disallowed to
call sys_exit()). Affected binaries include dhclient and uniq. The latter's
crashes cause obscure build failures.
This commit makes freebsd32 cap_enter() fail with [ENOSYS], as if capability
mode was not compiled in. Applications deal with this by doing their work
without capability mode.
This commit does not fix the uncommon situation where a 64-bit process
enters capability mode and then executes a 32-bit binary using fexecve().
This commit should be reverted when allowing the necessary freebsd32 system
calls in capability mode.
Since iconv was enabled (r254273, August 13), it has been impossible to
installworld using a read-only obj tree. This is common with NFS. Parts of
share/i18n unconditionally rebuild files like mapper.dir during
installation.
This patch ensures the files like mapper.dir are not rewritten with the same
contents.
Merge in support for PAPR-compliant (Power Architecture Platform
Requirements) systems from the projects/pseries branch. This in principle
includes all IBM POWER hardware released in the last 15 years with the
exception of POWER3-based systems when run in 64-bit mode. The main
development target, however, has been the PAPR logical partition support
that is the default target in KVM on POWER and QEMU -- mileage may vary
on actual hardware at present. Much of the heavy lifting here was done
by Andreas Tobler.
Neel Natu [Tue, 17 Sep 2013 16:06:07 +0000 (16:06 +0000)]
Fix a bug in decoding an instruction that has an SIB byte as well as an
immediate operand. The presence of an SIB byte in decoding the ModR/M field
would cause 'imm_bytes' to not be set to the correct value.
Fix this by initializing 'imm_bytes' independent of the ModR/M decoding.
Add a loader tunable to use only device tree-provided PCI devices. This is
needed on some more fragile systems to avoid machine checks when blindly
probing the PCI bus. Also reduce ofw_pcibus's priority slightly so that it
can be overridden.
Fix bug in busdma: if segs is a preexisting buffer, we memcpy it
into the DMA map. The length of the buffer had not yet been
initialized, however, so this would copy gibberish unless it
happened to be right by chance. This bug mostly only affected
systems with IOMMUs.
Zbigniew Bodek [Mon, 16 Sep 2013 10:39:35 +0000 (10:39 +0000)]
Implement pmap_advise() for ARMv6/v7 pmap module
Apply the given advice to the specified range of addresses within the
given pmap. Depending on the advice, clear the referenced and/or
modified flags in each mapping. Superpage within the given range will
be demoted or destroyed.
Reviewed by: alc
Approved by: cognet (mentor)
Approved by: re
Zbigniew Bodek [Mon, 16 Sep 2013 10:34:44 +0000 (10:34 +0000)]
Write protect base page after superpage demotion so that it may repromote
When clearing the modification status of the superpage, one of the
base pages produced during demotion should be marked as write disabled.
The intention is that subsequent write access may repromote.
In the current implementation this was done wrong as write permission was
granted instead of forbidden.
Remove zero-copy sockets code. It only worked for anonymous memory,
and the equivalent functionality is now provided by sendfile(2) over
posix shared memory filedescriptor.
Remove the cow member of struct vm_page, and rearrange the remaining
members. While there, make hold_count unsigned.
Requested and reviewed by: alc
Tested by: pho
Sponsored by: The FreeBSD Foundation
Approved by: re (delphij)
In pmap_copy(), when the copied region is mapped with superpage but does
not cover entire superpage, avoid copying. Doing partial copy would
require demotion, which is incompatible with the already held locks.
Reported by: cperciva
Reviewed by: alc
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Approved by: re (delphij)
Mark Johnston [Sun, 15 Sep 2013 21:38:46 +0000 (21:38 +0000)]
Use the address of the inpcb rather than the tcpcb to identify TCP
connections. This keeps the tcp provider consistent with the other network
providers.
Add a kernel interface (OF_xref_phandle()) for systems where phandles
used as cross-references in the device tree and phandles as used by the
Open Firmware client interface are in different namespaces. This include
IBM pSeries hardware as well as FDT systems. FDT certainly abuses
ihandles for this purpose and should be modified to use this API
eventually. This changes no behavior on systems where FreeBSD already
worked.
Reviewed by: marius
Approved by: re (kib)
MFC after: 2 weeks
The unbound-control-setup script needs to be generated so it knows where
to place the keys. Also, the correct umask is 027, not 026, although it's
not likely to make any difference.
Ed Schouten [Sat, 14 Sep 2013 22:09:10 +0000 (22:09 +0000)]
Properly remove c++filt.
The c++filt binary is only installed if ${MK_GCC} == yes && ${MK_CXX} ==
yes. This means that it should be removed if ${MK_GCC} == no ||
${MK_CXX} == no. In its current form, it actually uses a conjunction
instead of a disjunction.
As there is already a separate block for ${MK_CXX} == no listing
c++filt, simply remove the conditional entirely.
drm/radeon: Fix usage of pci_save_state() and pci_restore_state()
Calling those functions with the drmn device as argument causes a panic,
because it's not a direct child of pci$N. They must be called with the
vgapci device instead.
This fix is not enough to make suspend/resume work reliably.
vga_pci_(un)map_bios() takes a vgapci device as argument, not a drmn
one. This fixes a bug where the BIOS couldn't be mapped if the device
wasn't the boot display.
Approved by: re (kib; blanket for following drm2/radeon commits)
If the last page of the file is partially full and whole valid
portion is invalidated, invalidate the whole page. Otherwise,
partially valid page appears on a page queue, which is wrong. This
could only happen for the last page, because only then buffer which
triggered invalidation could not cover the whole page.
Reported and tested by: pho (previous version)
Reviewed by: alc
Sponsored by: The FreeBSD Foundation
Approved by: re (delphij)
MFC after: 2 weeks
add support for writing the pid of the daemon program to a pid file so
that daemon can be used w/ rc.subr and ports can use the additional
functionality, such as keeping the ldap daemon up and running, and have
the proper program to signal to exit..
PR: bin/181341
Submitted by: feld
Approved by: re (glebius)
When opening or closing fifo, ensure that the vnode is locked
exclusively. Filesystems are assumed to disable shared locking for
the fifo vnode locks, but some do not.
Reported and tested by: olgeni
Discussed with: avg
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Approved by: re (glebius)
Reduce the scope of the proctree_lock. If several processes cause
continuous calls to the uprintf(9), the proctree_lock could be
shared-locked for indefinite amount of time, starving exclusive
requests. Since proctree_lock is needed for fork() and exit(), this
effectively stops the machine.
While there, do the similar reduction for tprintf(9).
Reported and tested by: pho
Reviewed by: ed
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Approved by: re (glebius)
Change VM object lock assertion to match locking higher in the call
chain. This repairs a panic observed during pageout on some 64-bit
PowerPC systems.
Fix an issue that caused Integrated RAID volumes on LSI mps(4) controllers
to not get scanned on boot.
The problem originated in change 253549. With the change to the mps(4)
driver to scan only targets that it knows it has (as opposed to scanning
the entire bus), scanning RAID volumes on boot was omitted.
So, for versions of FreeBSD that have the scanning changes
(__FreeBSD_version 1000039 and higher), scan RAID volumes that are added
whether or not we're booting.
PR: kern/181784
Reported by: Xiguang Wang <kurapica@gmail.com>
Tested by: Dennis Glatting <dg@pki2.com>
Sponsored by: Spectra Logic
Approved by: re (delphij)
MFC After: 3 days
After r255321, clang uses libc++ by default. This leads to a lot of
errors when you enable WITH_GNUCXX to build libstdc++, since it will
include C++ headers from the libc++ installation under ${WORLDTMP}, and
those are not compatible with libstdc++ at all.
To fix this, add -stdlib=libstdc++ to CXXFLAGS when building libstdc++
(and its companion libsupc++) with clang.
John Baldwin [Thu, 12 Sep 2013 20:49:20 +0000 (20:49 +0000)]
- Document the UQUAD sysctl variants.
- Clarify that exactly one of the "access" flags is required and
list the optional flags in a separate list. Prefer bundling
CTLFLAG_TUN into the access flag by not documenting it as an
optional flag to set.
projects / FreeBSD / FreeBSD.git / log