Ed Schouten [Wed, 24 Aug 2016 11:35:49 +0000 (11:35 +0000)]
Add a Makefile for building the cloudabi32 kernel module.
Where the cloudabi64 kernel can be used to execute 64-bit CloudABI
binaries, this one should be used for 32-bit binaries. Right now it
works on i386 and amd64.
Ed Schouten [Wed, 24 Aug 2016 10:51:33 +0000 (10:51 +0000)]
Make execution of 32-bit CloudABI executables work on amd64.
A nice thing about requiring a vDSO is that it makes it incredibly easy
to provide full support for running 32-bit processes on 64-bit systems.
Instead of letting the kernel be responsible for composing/decomposing
64-bit arguments across multiple registers/stack slots, all of this can
now be done in the vDSO. This means that there is no need to provide
duplicate copies of certain system calls, like the sys_lseek() and
freebsd32_lseek() we have for COMPAT_FREEBSD32.
This change imports a new vDSO from the CloudABI repository that has
automatically generated code in it that copies system call arguments
into a buffer, padding them to eight bytes and zero-extending any
pointers/size_t arguments. After returning from the kernel, it does the
inverse: extracting return values, in the process truncating
pointers/size_t values to 32 bits.
Ed Schouten [Wed, 24 Aug 2016 10:36:52 +0000 (10:36 +0000)]
Remove an unused header file.
The native CloudABI data types header file used to be pulled in by the
vDSOs when they were still written in C. Since they are now all
rewritten in assembly, this can go away.
Ed Schouten [Wed, 24 Aug 2016 10:13:18 +0000 (10:13 +0000)]
Convert pointers obtained from the threadattr_t structure with TO_PTR().
In all of these source files, the userspace pointer size corresponds
with the kernelspace pointer size, meaning that casting directly works.
As I'm planning on making 32-bit execution on 64-bit systems work as
well, use TO_PTR() here as well, so that the changes between source
files remain minimal.
Julio Merino [Wed, 24 Aug 2016 10:10:26 +0000 (10:10 +0000)]
Skip ls tests that use sparse files if these are not supported.
Some of the ls(1) tests create really large sparse files to validate
the number formatting features of ls(1). Unfortunately, those tests fail
if the underlying test file system does not support sparse files, as is the
case when /tmp is mounted on tmpfs.
Before running these tests, check if the test file system supports sparse
files by using getconf(1) and skip them if not. Note that the support for
this query was just added to getconf(1) in r304694.
Michael Tuexen [Wed, 24 Aug 2016 06:22:53 +0000 (06:22 +0000)]
When aborting an association, send the ABORT before notifying the upper
layer. For the kernel this doesn't matter, for the userland stack, it does.
While there, silence a clang warning when compiling it in userland.
Bruce Evans [Wed, 24 Aug 2016 05:54:11 +0000 (05:54 +0000)]
Fix key delay and repeat, part 2.
Use sbintime_t timeouts with precision control to get very accurate
timing. It costs little to always ask for about 1% accuracy, and the
not so new event timer implementation usual delivers that, and when
it can't it gets much closer than our previous coarse timeouts and
buggy simple countdown.
The 2 fastest atkbd repeat rates have periods 34 and 38 msec, and ukbd
pretended to support rates in between these. This requires
sub-microsecond precision and accuracy even to handle the 4 msec
difference very well, but ukbd asked the timeout subsystem for timeouts
of 25 msec and the buggy simple countdown of this gave a a wide range
of precisions and accuracies depending on HZ and other timer
configuration (sometimes better than 25 msec but usually more like 50
msec). We now ask for and usually get precision and accuracy of about
1% for each repeat and much better on average.
The 1% accuracy is overkill. Rounding of 30 cps to 34 msec instead of
33 already gives an error of +2% instead of -1%, and ut AT keyboards on
PS/2 interfaces have similar errors.
A timeout is now scheduled for every keypress and release. This allows
some simplifications that are not done. It allows removing the timeout
scheduling for exiting polled mode where it was unsafe in ddb mode. This
is done. Exiting polled mode had some problems with extra repeats. Now
exiting polled mode lets an extra timeout fire and the state is fudged
so that the timeout handler does very little.
The sc->time_ms variable is unsigned to avoid overflow. Differences of
it need to be signed. Signed comparisons were emulated by testing an
emulated sign bits. This only works easily for '<' comparisonss, but
we now need a '<=' comparison. Change the difference variable to
signed and use a signed comparison. Using unsigned types here didn't
prevent overflow bugs but just reduced them. Overflow occurs with
n repeats at the silly repeat period of [U]INT_MAX / n. The old countdown
had an off by 1 error, and the simplifications would simply count down
1 to 0 and not need to accumulate possibly-large repeat repeats.
Sepherosa Ziehau [Wed, 24 Aug 2016 04:21:15 +0000 (04:21 +0000)]
hyperv/hn: Log a warning for RESET_CMPLT.
RESET is not used by the hn(4) at all, and RESET_CMPLT does not even
have a rid to match with the pending requests. So, let's put it
onto an independent switch branch and log a warning about it.
MFC after: 1 week
Sponsored by: Microsoft
Differential Revision: https://reviews.freebsd.org/D7602
Justin Hibbits [Wed, 24 Aug 2016 03:51:40 +0000 (03:51 +0000)]
Fix system hang when large FDT is in use
Summary:
Kernel maps only one page of FDT. When FDT is more than one page in size, data
TLB miss occurs on memmove() when FDT is moved to kernel storage
(sys/powerpc/booke/booke_machdep.c, booke_init())
This introduces a pmap_early_io_unmap() to complement pmap_early_io_map(), which
can be used for any early I/O mapping, but currently is only used when mapping
the fdt.
Submitted by: Ivan Krivonos <int0dster_gmail.com>
Differential Revision: https://reviews.freebsd.org/D7605
Mike Karels [Wed, 24 Aug 2016 00:52:30 +0000 (00:52 +0000)]
Fix L2 caching for UDP over IPv6
ip6_output() was missing cache invalidation code analougous to
ip_output.c. r304545 disabled L2 caching for UDP/IPv6 as a workaround.
This change adds the missing cache invalidation code and reverts
r304545.
Emmanuel Vadot [Tue, 23 Aug 2016 22:26:50 +0000 (22:26 +0000)]
Allwinner: Add thermal sensor driver for A10/A20
The thermal sensor lives in the touch screen controller. Touch screen part
isn't done for now.
Temperature is read every ~2 seconds and exposed via sysctl.
LC_*_MASK bit shifting order was partially broken from the initial commit
time at year 2012. Only LC_COLLATE_MASK and LC_CTYPE_MASK are in the
right order.
The order here should match XLC_* from "xlocale_private.h" which, in turn,
match LC_* publicly visible order from <locale.h> which determines how
locale components are stored in the structure.
LC_*_MASK -> XLC_* translation done as "ffs(mask) - 1" in the querylocale()
and equivalent shift loop in the newlocale(), so mapped to some wrong
components (excluding two mentioned above).
Formally the fix is ABI breakage, but old code using those masks
never works properly in any case.
Only newlocale() and querylocale() are affected.
Gleb Smirnoff [Tue, 23 Aug 2016 20:04:23 +0000 (20:04 +0000)]
The -f check here is used to determine whether we have a single kernel
config or a list of them. Put the variable into quotes, to avoid syntax
error from [ in case of list. Without this change list is still working,
but an error is reported in the build log file.
Bruce Evans [Tue, 23 Aug 2016 19:50:16 +0000 (19:50 +0000)]
Fix key delay and repeat, part 1.
kbdcontrol -r fast is documented to give a non-emulated atkbd's fastest
rate of 250.34, but is misimplemented to request this as 0.0. ukbd
supports many nonstandard rates, although it is currently too inaccurate
by a factor of several hundred for non-huge nonstandard rates to be
useful. It mapped 0.0 to 200.0. A repeat delay of 0 means a rate of
infinity which is quite fast, but physical constraints limit this to
a few MHz and the inaccuracies made it almost usable.
Convert 0.0 to the documented 250.34.
Also convert negative args and small args to the 250.34 minimal ones,
like atkbd does. This is for KDSETREPEAT -- the 2 versions of the
deprecated KDSETRAD have bounds checking. Keep not doing any bounds
checking or conversions for upper limits since nonstandard large
delays are useful for testing.
The inaccuracies are dependent on HZ and the timeout implementation.
With the old timeout implementation and HZ = 1000, 200.0 probably
worked better to emulate 250.34 than 250.34 itself. HZ = 100 gives
roundoff errors that accidentally reduce the inaaccuracies, and
event timers reduce the inaccuracies even more, so 200.0 was giving
more like itself (perhaps 215.15 on average but sometimes close to
10 msec repeat which is noticebly too fast). This commit makes 0.0
noticeably too slow, like 250.34 always was.
Bryan Drewery [Tue, 23 Aug 2016 19:37:18 +0000 (19:37 +0000)]
FAST_DEPEND: Fix 'make all install' not properly rebuilding based on .depend.* files.
An optimization is in place to skip reading the .depend.* files with
'make install'. This was too strong and broke 'make all install' and
'make foo.o foo install'. Now only skip reading the dependency files
if all make targets ran are install targets.
The problem comes about because headers are only added in as a guessed
dependency if .depend.* files do not yet exist. If they do exist, even
if being skipped from being read, then the header dependencies are not
applied. This applies to all #included files, and not just headers.
Dimitry Andric [Tue, 23 Aug 2016 19:31:43 +0000 (19:31 +0000)]
In addition to creating subdirectories under .OBJDIR for SRCS with
relative paths, also create them for DPSRCS. This is needed for builds
that generate files during the depend stage, which cannot be compiled by
themselves, since those have to be put in DPSRCS.
Bryan Drewery [Tue, 23 Aug 2016 19:29:37 +0000 (19:29 +0000)]
Fix in-tree GCC builds after r304681.
There were a few issues.
- In-tree GCC won't have X_COMPILER_TYPE defined but will have
WANT_COMPILER_TYPE==gcc set from the SYSTEM_COMPILER logic that can
be used. Make the clang check specific to clang as well to ensure
-target doesn't leak into a GCC build.
- When using a cross-compiler GCC (with a default sysroot or arch) and also
passing --sysroot, it basically forgets all internal paths for
libraries. We've already worked around this quite a bit for
the external toolchains. Now for the in-tree bootstrap cross-compiler
GCC, also pass in the needed -B${WORLDTMP}/usr/lib to find the crt
object files, but also -isystem and -L to fix the paths. This creates
quite a spammy build log, but it is clear and still achieves the goals
and stays consistent between internal and external build flags.
Reducing the spam by using the '=' prefix feature will help and be
done later.
MFC after: 3 days
X-MFC-With: r304681
Reported by: bz
Pointyhat to: bdrewery
Sponsored by: EMC / Isilon Storage Division
Landon J. Fuller [Tue, 23 Aug 2016 19:03:11 +0000 (19:03 +0000)]
bhndb(4): Fix unsigned integer underflow in dynamic register window
handling. This resulted in the window target being left uninitialized
when an underflow occured.
Approved by: adrian (mentor)
Differential Revision: https://reviews.freebsd.org/D7617
Andrew Turner [Tue, 23 Aug 2016 15:48:27 +0000 (15:48 +0000)]
Include the offset the virtual address is within an L1 or L2 block when
finding the vm_page_t in pmap_extract_and_hold. Previously it would return
the vm_page_t of the first page in a block. This would cause issues when,
for example, fsck reads from a device into the middle of a superpage. In
this case the read call would write to the start of the block, and not to
the buffer passed in.
Obtained from: ABT Systems Ltd
MFC after: 1 month
Sponsored by: The FreeBSD Foundation
Bryan Drewery [Tue, 23 Aug 2016 15:22:17 +0000 (15:22 +0000)]
Re-enable WITH_SYSTEM_COMPILER with WITH_META_MODE.
This was disabled in r301468 due to -target/--sysroot sometimes being used in
the build and other times not being used. Now that it is always used since
r304681, it is safe to combine the features.
Bryan Drewery [Tue, 23 Aug 2016 15:20:32 +0000 (15:20 +0000)]
Always pass in -target and --sysroot flags for the build.
The internal bootstrap compiler has a default sysroot set by TOOLS_PREFIX
and target set by TARGET/TARGET_ARCH. However, there are several needs to
always pass an explicit --sysroot and -target.
- External compiler needs sysroot and target flags.
- External ld needs sysroot.
- To be clear about the use of a sysroot when using the internal compiler.
- Easier debugging.
- Allowing WITH_SYSTEM_COMPILER+WITH_META_MODE to work together due to
the flip-flopping build command when sometimes using external and
sometimes using internal.
- Allow using no lld which has support for default paths.
The default sysroot in the bootstrap compiler is not changed. The
buildenv compiler will still work with its default and will also
include -target/--sysroot from CC in the environment.
Andriy Gapon [Tue, 23 Aug 2016 10:40:53 +0000 (10:40 +0000)]
intpm: add support for SB800
This code should be able to support later AMD chipsets as well, but that
hasn't been tested.
SB800 supports accessing several different SMBus buses using the same
set of constrol registeirs plus special PMIO registers that control which
bus is selected. This could be exposed to consumers as several smb devices
each talking to its bus. This feature is not implemented yet.
Justin Hibbits [Tue, 23 Aug 2016 04:37:03 +0000 (04:37 +0000)]
tlb1_init() can be called twice on BookE
Summary:
There is no need to call tlb1_init() twice. Now it is called first time from
booke_init() and second time from powerpc_init() (where it is under BOOKE
switch). Although this does not cause immediate problems in the mainline kernel,
this can lead to undesirable side effects like two TLB entries with the same VA
in the TLB1. Presence of two TLB entries with the same VA can hang CPU.
Test Plan:
Add initial mapping for UART to the tlb1_init(), build and boot the kernel,
ensure that mapping presents only once (most convinient way - through Lauterbah
or similar hardware debugger)
Submitted by: Ivan Krivonos <int0dster_gmail.com>
Differential Revision: https://reviews.freebsd.org/D7607
Pedro F. Giffuni [Tue, 23 Aug 2016 02:07:08 +0000 (02:07 +0000)]
indent(1): Fix off-by-one in control flow leading dead code.
Coverity correctly reported that it's impossible for /comparison/ to be 0
here, because the only way for the for loop to end is by /comparison/
being < 0.
Fortunately the consequences of this bug weren't severe; for duplicated
entries in the typedef names file it would unnecessarily duplicate strings
with strdup(), but pointers to those would replace existing ones. So this
was a memory leak at worst.
Pedro F. Giffuni [Tue, 23 Aug 2016 01:40:45 +0000 (01:40 +0000)]
indent(1): Fix memory leaks pointed out by clang-analyzer.
Shift the responsibility of allocating memory for the string duplicate
from the caller (set_option, add_typedefs_from_file) to the callee
(add_typename) as it has more knowledge about when the duplication
actually needs to occur.
Bryan Drewery [Mon, 22 Aug 2016 22:51:07 +0000 (22:51 +0000)]
For 'make <directory>' hook into the all_subdir_<directory> targets.
This fixes parallel build issues when trying to depend on ${SUBDIR}. An
example of this in share/i18n/csmapper/Makefile where mapper.dir depends
on ${SUBDIR} having been traversed and built already. Before this
change running make in that directory would build the subdirectories
twice. This led to obscure build races. While reworking that build
may be possible, the framework should not so easily allow creating such
problems.
Now depending on <directory> will properly redirect to the
all_subdir_<directory> target rather than invoking the inline shell.
This also makes 'make -jX <directory>' now respect any
SUBDIR_DEPEND_<directory> statements when SUBDIR_PARALLEL is defined.
This is not entirely intended and may be changed later.
Bryan Drewery [Mon, 22 Aug 2016 22:51:01 +0000 (22:51 +0000)]
Stop using _SUBDIR internally for non-SUBDIR_PARALLEL builds.
This is unifying more of the logic. Rather than create targets such
as 'all: all_subdir_foo' when using SUBDIR_PARALLEL and using
'all: _SUBDIR' when not using SUBDIR_PARALLEL, always use the
expanded out <target>_subdir_<directory> pattern. When not using
SUBDIR_PARALLEL, have each directory-target depend on the previously
defined targets as to respect the *order* of SUBDIR.
Using 'make -N' now prints all directory traversals individually rather
than using a loop, since a loop is no longer used to traverse.
This is part of an effort to cleanup handling of some edge cases
involving 'make <directory>' and making it simpler in the sense
that the pattern used to build is the same for all modes.
Bryan Drewery [Mon, 22 Aug 2016 22:50:58 +0000 (22:50 +0000)]
Fix building on read-only source trees.
This partially reverts r296702 and reworks the original check to only
look in .CURDIR. This avoids ever trying to rebuild a .src file that is
already in the source tree as an override.
1) Back out r304607 case 2). fgetwln() as its pair fgetln() supposed to
return partial line on any errors. See the comment in fgetln.c.
Add corresponding comment to fgetwln() too.
2) Rewrite r304607 case 1).
3) Remove "Fast path" from __fgetwc_mbs() since it can't detect encoding
errors and ignores them all.
Don't separate the status stage of the XHCI USB control transfers into
its own job because this breaks the simplified QEMU XHCI TRB parser,
which expects the complete USB control transfer as a series of back to
back TRBs. The old behaviour is kept under #ifdef in case this change
breaks enumeration of any USB devices.
Ed Schouten [Mon, 22 Aug 2016 17:37:31 +0000 (17:37 +0000)]
Make CloudABI work on i386.
Copy over amd64's cloudabi64_sysvec.c into i386 and tailor it to work.
Again, we use a system call convention similar to FreeBSD, except that
there is no support for indirect system calls (%eax == 0).
Where i386 differs from amd64 is that we have to store thread/process
entry arguments on the stack instead of using registers. We also have to
put an extra pointer on the stack for TLS (for GSBASE). Place that
pointer in the empty slot that is normally used to hold return
addresses. That seems to keep the code simple.
Bruce Evans [Mon, 22 Aug 2016 16:39:51 +0000 (16:39 +0000)]
Further fixes for translation of PrintScreen/SysRq.
The previous fix was tested mainly on 3 AT keyboards with USB adaptors where
it works. 1 USB keyboard doesn't translate Alt-PrintScreen, so the software
has to do it.
Reorganize a little to share some code and to not translate the unusual usb
scan code0x8a unless an Alt modified is set. Remove redundant check of Alt
modifiers. Translation now more clearly filters out Alt-PrintScreen before
the check.
The table of errors fixed in the previous commit had many bugs. Correct
table:
K_RAW Ctl-PrintScreen: E0-2A-E0-37 -> E0-37
K_RAW Alt-PrintScreen (with 4 comb. of Ctl/Shift): 79 -> 54
K_RAW Pause/Break (with 4 comb. of Alt/Shift): E0-46 -> E1-1D-45
K_CODE PrintScreen (with 4 comb. of Ctl/Shift): 54 -> 5c
K_CODE Alt-PrintScreen (with 4 comb. of Ctl/Shift): 7e -> 54
K_CODE Pause/Break (with 8 comb. of Ctl/Alt/Shift): 6c -> 68
That is 25 of 32 shift combinations for 2 keys fixed. All 16 combinations
were broken for K_CODE and thus also for K_XLATE.
Fix error processing.
1) Don't forget to set __SERR on __slbexpand() error.
2) Check for __fgetwc() errors using errno. Don't check for __SERR
as PR suggested, it user-visible flag which can stick from previous
functions and stdio code don't check it for this purpose.
Ryan Stone [Mon, 22 Aug 2016 15:27:37 +0000 (15:27 +0000)]
Temporarily disable the optimization from r304436
r304436 attempted to optimize the handling of incoming UDP packet by only
making an expensive call to in_broadcast() if the mbuf was marked as an
broadcast packet. Unfortunately, this cannot work in the case of point-to-
point L2 protocols like PPP, which have no notion of "broadcast".
Discussions on how to properly fix r304436 are ongoing, but in the meantime
disable the optimization to ensure that no existing network setups are broken.
Conrad Meyer [Mon, 22 Aug 2016 14:51:09 +0000 (14:51 +0000)]
ioat(4): Allow callouts to be scheduled after hw reset
is_completion_pending governs whether or not a callout will be scheduled
when new work is queued on the IOAT device. If true, a callout is
already scheduled, so we do not need a new one. If false, we schedule
one and set it true. Because resetting the hardware completed all
outstanding work but failed to clear is_completion_pending, no new
callout could be scheduled after a reset with pending work.
This resulted in a driver hang for polled-only work.
Fix for invalid use of bits in input context. Basically split
configuring of EP0 and non-EP0 into xhci_cmd_evaluate_ctx() and
xhci_cmd_configure_ep() respectivly. This resolves some errors when
using XHCI under QEMU and gets is more in line with the XHCI
specification.
Sepherosa Ziehau [Mon, 22 Aug 2016 07:34:39 +0000 (07:34 +0000)]
hyperv/hn: Move chimney sending buffer to hn_softc
And don't recreate chimney sending buffer for each primary channel
open, it is now created in device_attach DEVMETHOD and destroyed
in device_detach DEVMETHOD.
MFC after: 1 week
Sponsored by: Microsoft
Differential Revision: https://reviews.freebsd.org/D7574
Michael Tuexen [Mon, 22 Aug 2016 01:45:29 +0000 (01:45 +0000)]
Improve the locking when sending user messages.
First, keep a ref count on the stcb after looking it up, as
done in the other lookup cases.
Second, before looking again at sp, ensure that it is not
freed, because the assoc is about to be freed.
Pyun YongHyeon [Mon, 22 Aug 2016 01:19:05 +0000 (01:19 +0000)]
Add Killer E2400 Gigabit Ethernet support.
It seems Killer E2200/E2400 has a BIOS misconfiguration or silicon
bug which triggers DMA write errors when driver uses advertised
maximum payload size. Force the maximum payload size to 128 bytes
in DMA configuration.
This change should fix occasional DMA write errors reported on
Killer E2200.