Commit Graph

265192 Commits

Author SHA1 Message Date
mrg 55fccfafc7 update status.
status is:  with uncommited changes, some ports (amd64, sparc) work.
2019-02-02 12:09:55 +00:00
cherry 6ac6119272 Summary: Disable ixg(4) until the linuxist mb() API re-definition
within the XEN amd64 dom0 build is resolved.

Detail:
The xen public headers >= 0x00030201 depend on a linuxism - a memory
barrier "API" namely mb(), rmb() and wmb(). These are made accessible
to NetBSD code via the interface
sys/external/bsd/common/include/asm/barrier.h

Currently the ixg(4) driver uses an identical API by defining them
(incorrectly) to inline assembler macro definitions in
sys/dev/pci/ixgbe/ixgbe_osdep.h

This will result in the dom0 builds failing to build due to this
redefinition of the macros. We therefore disable the driver in
NetBSD/amd64 dom0 in order to facilitate smoothly moving the XEN
kernel builds to the new XEN source level API (RELEASE-4.11.1).

The fix for this is trivial, but needs the attention of the driver
maintainer.
2019-02-02 12:05:09 +00:00
cherry a215bf0172 Import Xen public headers version 4.11.1
Release tag: RELEASE-4-11-1

git log of the freeze is as follows:

commit 96cbd0893f783997caaf117e897d5fa8f2dc7b5f
Author: Jan Beulich <jbeulich@suse.com>
Date:   Thu Nov 29 15:04:11 2018 +0100

    update Xen version to 4.11.1
2019-02-02 08:03:47 +00:00
isaki c23eabedab Correct debug messages. 2019-02-02 05:02:03 +00:00
isaki 8b1980912d Fix minor bugs of SNDCTL_DSP_GETISPACE.
- hiwat is playback-only parameter.
- 'bytes' should not be rounded down.
2019-02-02 04:52:16 +00:00
mrg 3eb0897352 regen for earmv4 and GCC 7. 2019-02-02 01:13:18 +00:00
sevan 166b2d1655 Add MFU 2019-02-01 21:23:22 +00:00
mrg b9c25d2c43 fix merge botch. 2019-02-01 13:12:22 +00:00
mrg 7d798d96e5 regen mknative for ppc and sparc. 2019-02-01 12:49:21 +00:00
maxv 6857c48bf0 Add the remaining pmap callbacks, will be used by NVMM-VMX. 2019-02-01 11:35:13 +00:00
mrg 0f7ab0c2e3 - update source lists
- XXX: pull out OBJS-libcommon-target and OBJS-libcommon-target
  in mknative
2019-02-01 10:39:48 +00:00
mrg ce4fd745ec use -Wno-error=maybe-uninitialized for one of the new ubsan files. 2019-02-01 10:39:03 +00:00
mrg 44a4653fe8 regen 2019-02-01 10:37:38 +00:00
mrg 75b1676ff7 - new shlib minor
- move -std=* handling, besides the list in CXX11_ALWAYS, to
  libsupc++/Makefile.common
- include the arch defs.mk to obtain ${G_*_SOURCES} and use them
  for -std= setting
2019-02-01 10:37:13 +00:00
mrg 437aa6a459 - use ADD_FRONTEND_LIBRARY to avoid linking libfrontend
- add a hack to ensure driver-i386.c gets built before linking
2019-02-01 10:36:33 +00:00
mrg ed4bd2e275 - be sure bsd.own.mk has been included 2019-02-01 10:35:09 +00:00
mrg 88241950de - move various stuff into Makefile.prog
- make sure VER_CPPFLAGS is applied to all required files
- add ADD_FRONTEND_LIBRARY to avoid linking libfrontend, while
  using the rest of this fragment
2019-02-01 10:34:18 +00:00
mrg ccf5bd012a - move driver file config to Makefile.driver. 2019-02-01 10:32:32 +00:00
mrg 7c87265e53 - new shlib majors 2019-02-01 10:31:55 +00:00
mrg b42a39b7de - use --with-default-libstdcxx-abi=new for GCC, otherwise we get the
old ABI with --disable-libstdcxx-dual-abi
- fix libbacktrace mknative
- find gthr-default.h for libstdc++ mknative
- fetch FILESYSTEM_SOURCES
2019-02-01 09:27:23 +00:00
mrg 8dec05958d don't check the return value of execvp() for failure to call err().
assume if it returns at all something has failed.
2019-02-01 09:06:07 +00:00
mrg 2235c7e9bc make a fallthru comment GCC7 compatible. 2019-02-01 08:37:21 +00:00
mrg cdf28a3078 fix FALLTRHOUGH spell-o. 2019-02-01 08:33:29 +00:00
mrg 684b182f81 compare pointers with NULL not '\0'. 2019-02-01 08:29:03 +00:00
maxv 2089a3819a Fix two issues:
* Uh I put the wrong masks in some GPRs, fuck.

 * When the opsize of MOVZX is 4, we need to combine the zero-extend from
   the instruction with the natural zero-extend of long mode.

Add two associated tests.
2019-02-01 06:49:58 +00:00
maxv b1bb495249 Change the format of the pp_attrs field: instead of using PTE bits
directly, use abstracted bits that are converted from/to PTE bits when
needed (in pmap_sync_pv).

This allows us to use the same pp_attrs for pmaps that have PTE bits at
different locations.
2019-02-01 05:44:29 +00:00
maxv a8470e3880 Put correct values in the seg fields. AMD doesn't check for that, but Intel
does, so they need to be correct.
2019-02-01 05:32:08 +00:00
christos 93f9738b24 Add a software beacon.
XXX: As all the other usb software beacons this is driven at fixed intervals
when usb task fires.
2019-02-01 03:20:35 +00:00
pgoyette 8ca245f5b0 Now that the kernel/module interface has stabilized after the recent
merge of the [pgoyette-compat] branch, update the kernel version to
reflect all of the interim changes.

Welcome to 8.99.33 !!!

OK mrg@
2019-01-31 22:07:46 +00:00
mrg e27558b901 remove our riscv changes in gcc.old. they were obsolete with GCC 5
update, let alone GCC 6.  new GCC 7 has native support again.

this is really about trying to avoid conflicts when we next have
gcc become gcc.old.
2019-01-31 20:44:46 +00:00
maxv 5fd9c1560b Move some code into a separate function, no functional change. 2019-01-31 20:42:31 +00:00
maxv ca1f7519df Fix kernel info leaks. 2019-01-31 20:09:05 +00:00
jakllsch db17a1610e Reset the hardware like other drivers do for better compatibility with
more chip firmware versions.

From (and should fix) PR kern/53913.
2019-01-31 18:21:21 +00:00
christos 909cb6d3de remove unused variable 2019-01-31 16:03:50 +00:00
msaitoh c0a7c67136 Add yet another PHY workaround. Disable generation of early preamble on 82577
PHY rev. 1 or 2 and 82578 PHY rev. 1.
2019-01-31 15:30:23 +00:00
rin fc9aba3436 Correct error bits in RX packet header:
- Drop error is bit 31.
- CRC error is bit 29.
- Distinguish the two errors in debug printf.
2019-01-31 15:27:57 +00:00
rin ed82019e65 Enable hash-table-based multicast filter:
- Drop AXEN_RXCTL_ACPT_PHY_MCAST bit (neither Linux nor FreeBSD sets it).
- 0x0010 bit is not AXEN_RXCTL_ACPT_BCAST ("accept broadcast") but
  AXEN_RXCTL_ACPT_MCAST ("accept multicast (hash-table-based)").
2019-01-31 15:26:24 +00:00
rin 3c9aa36807 In axen_tick_task(), call axen_miibus_statchg() when link is lost
(like other drivers). Otherwise, link can be lost forever, which
results in sysmtem freeze.
2019-01-31 15:24:13 +00:00
rin 4fc0ae6b33 Enable AXEN_RXCTL_START bit only when RX is ready. Otherwise,
the adapter eventually falls into "no carrier" state while it
is not running.
2019-01-31 15:22:56 +00:00
rin fd382a491f Fix previous:
- axen_coe() should be called from axen_init(), not axen_ax88179_init().
- Simplify logic in axen_coe(); AXEN_[RT]XCOE_OFF = 0.
2019-01-31 15:21:05 +00:00
skrll ff397156cf Only free the _init_memory pages if all APs start corectly. 2019-01-31 13:26:21 +00:00
skrll 0a47c48a44 Fix another build 2019-01-31 13:19:19 +00:00
skrll 3bf2ce81ff Fix build 2019-01-31 13:16:31 +00:00
skrll a476a90d44 Change ap_mpstart to return non-zero value if any/all APs don't start. 2019-01-31 13:06:10 +00:00
christos 653e228141 Prevent integer overflow: ioctl commands are u_long.
For consistency move prototype to the _mod.h header.

XXX: Why are the compat files here? Shouldn't they be in compat/common?
Or because this could be a separate module, they belong with it?
2019-01-31 12:31:50 +00:00
msaitoh 900e7dd5ab Stop continuing /var/shm check when var_shm_symlink is in /etc/rc.conf.
OK'd by martin.
2019-01-31 10:06:32 +00:00
mrg 75c0f414f8 revert these two changes which cause various parts of libstdc++ to fail
to build with GCC7:

revision 1.3
date: 2017-07-22 13:52:52 -0700;  author: joerg;  state: Exp;  lines: +2 -1;  commitid: gesFYL8PorhYCg0A;
One more missing check for DECL_INITIAL being non-NULL.

revision 1.2
date: 2017-07-17 12:53:10 -0700;  author: joerg;  state: Exp;  lines: +2149 -1573;  commitid: fxGaJg3EuIcnsCZz;
branches:  1.2.2;
A const declaration with explicit section attribute should create a
read-only section, whether it is initialized or not.


XXX:  i may have merged them wrongly into gcc7, but AFAICT, they're
expecting to use named BSS sections, so this seems wrong now anyway.


joerg, please feel free to check :-)
2019-01-31 08:53:07 +00:00
mrg 92869bcb74 fix a bunch of merge botches. libasan now builds. 2019-01-31 08:44:14 +00:00
mrg ee5ee6ce23 only a relatively small number of files that aren't in the c++98/ dir
need to be compiled in gnu++98 mode.  list them explicitly instead of
assuming anything with no -std should be gnu++98.
2019-01-31 08:42:03 +00:00
mrg fffd49d1e5 - remove stale comment
- remove i386 a.out support
- use ${nbsd_tm_file} for all netbsd targets
- remove duplicated (old) riscv stuff
2019-01-31 06:59:31 +00:00