Summary:
This review ports mlx5 driver, kernel's OFED stack (userland is already enabled), KTLS and krping to powerpc64 and powerpc64le.
krping requires a small change since it uses assembly for amd64 / i386.
NOTE: On powerpc64le RDMA works fine in the userspace with libmlx5, but on powerpc64 it does not. The problem is that contrib/ofed/libmlx5/doorbell.h checks for SIZEOF_LONG but this macro exists on neither powerpc64* nor amd64. Thus, the file silently goes to the fallback function written for 32-bit architectures. It works fine on little-endian architectures, but causes a hard fail on big-endian. It's possible it may also cause some runtime issues on little-endian.
Thus, on powerpc64 I verified that RDMA works with krping.
Reviewers: #powerpc, hselasky
Subscribers: bdrewery, imp, emaste, jhibbits
Differential Revision: https://reviews.freebsd.org/D38786
When doing a fully incremental build (with WITHOUT_CLEAN enabled), from
a commit before llvm 15 was merged (3264f6b88f), to a commit after
that, a number of .inc files were not regenerated. This could lead to
unexpected compilation errors when these .inc files were included from
llvm-project sources, similar to:
In file included from /usr/src/contrib/llvm-project/clang/lib/CodeGen/CGBuiltin.cpp:8268:
/usr/obj/usr/src/amd64.amd64/lib/clang/libclang/clang/Basic/arm_mve_builtin_cg.inc:279:18: error: no matching constructor for initialization of 'clang::CodeGen::Address'
Address Val2 = Address(Val1, CharUnits::fromQuantity(2));
^ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Work around this by making the .inc files dependent on the tblgen binary
used for generating them. E.g., we can relatively safely assume that if
the binary gets updated, the .inc files must also be updated. (Although
this is not 100% optimal, the gain by complicating things even more is
probaby not worth the effort.)
MFC after: 3 days
Reviewed by: emaste
Differential Revision: https://reviews.freebsd.org/D38770
According to the Linux sources the kernel exposes a proces virtual
memory size via proc filesystem into the three files - stat, status
and statm. This is the struct mm->total_vm value adjusted to the
corresponding units - bytes, kilobytes and pages.
The fix is based on a fernape@ analysis.
PR: 265937
Reported by: Ray Bellis
MFC after: 3 days
- skipping end of block that is not open: Oc
- no blank before trailing delimiter
- remove useless TN macros
- remove commented out reference for esis(4)
MFC after: 5 days
Differential Revision: https://reviews.freebsd.org/D38783
Since the kernel options ROUTE_MPATH and FIB_ALGO are enabled
per default for a while, it's good to have some user facing
documetation about the general functionality of multipath
routing and fib lookup algorithms.
Reviewed by: pauamma, Jose Luis Duran <jlduran at gmail dot com>
MFC after: 5 days
Differential Revision: https://reviews.freebsd.org/D38783
Summary:
This review ports mlx5 driver, kernel's OFED stack (userland is already enabled), KTLS and krping to powerpc64 and powerpc64le.
krping requires a small change since it uses assembly for amd64 / i386.
NOTE: On powerpc64le RDMA works fine in the userspace with libmlx5, but on powerpc64 it does not. The problem is that contrib/ofed/libmlx5/doorbell.h checks for SIZEOF_LONG but this macro exists on neither powerpc64* nor amd64. Thus, the file silently goes to the fallback function written for 32-bit architectures. It works fine on little-endian architectures, but causes a hard fail on big-endian. It's possible it may also cause some runtime issues on little-endian.
Thus, on powerpc64 I verified that RDMA works with krping.
Reviewers: #powerpc, hselasky
Subscribers: bdrewery, imp, emaste, jhibbits
Differential Revision: https://reviews.freebsd.org/D38786
Bruce added stripping of gcc2_compiled and other symbols when he made
the boot loader load the symbols for the kernel in 1995 (b5d89ca8ad)
before the FreeBSD 2.1 release. This was copied around a bit and
tweaked over the years, but these symbols aren't produced by clang, nor
gcc12. The were to support dbx for a.out stabs format. gcc removed them
with stabs support last year. gcc 2.95.4 in FreeBSD 4.x continued to
emit these symbols unconditionally (it was missing a test for aout vs
elf it would appaer). They disappeared entirely with gcc 3.2.4 in 5.x
for all non a.out builds, and entirely in FreeBSD 6.x which had gcc
3.2.6.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D38764
Previously strtol was used and the result was directly cast to an int
without checking for an overflow. Use strtonum instead since it is
safer and tells us what went wrong.
Reviewed by: imp
Pull Request: https://github.com/freebsd/freebsd-src/pull/643
If a CPU for some reason returns 0 as CPU frequency, we currently panic
on the resulting divide by zero when trying to initialize the CPU(s) via
APIC. When this happens, we'll fallback to measuring the frequency
instead.
PR: 269767
Reviewed by: imp
Pull Request: https://github.com/freebsd/freebsd-src/pull/664
This is a variant of CK_*_FOREACH from FreeBSD queue.h which starts
iteration at the specified item. If the item pointer is NULL, iteration
starts from the beginning of the list.
Upstream commit 74366be35a6f4635f248a3c62d2d23245a4eb0f4.
MFC after: 2 weeks
Sponsored by: Klara, Inc.
Most options in kernel config files use "options<space><tab>OPTION".
This allows the option to be commented out without shifting columns.
A few options had two tabs, and some had spaces. Make them consistent.
I missed this file on the last pass.
- Be consistent with RFC references, so add a space after 'RFC'
- Add a LIBRARY section
- Use standard integer types in the SYNOPSIS section
Obtained from: DragonflyBSD
MFC after: 5 days
Differential Revision: https://reviews.freebsd.org/D27548
Summary:
The LTO issue has been fixed. While -flto for some reason is commented out,
since it wasn't completely removed, it may be expected to be reenabled.
Reviewers: se
Approved by: se
MFC after: 3 days
Subscribers: imp
Differential Revision: https://reviews.freebsd.org/D38755
This version adds a command to dc to query whether extended registers
are enabled or not.
(cherry picked from commit 61e1a12bb6c3bfdb0a4e499c88e8eaa2b548e427)
The routine is used as a general event-limiting routine in places which
have nothing to do with packets.
Provide a define to keep everything happy.
Reviewed by: rew
Sponsored by: Rubicon Communications, LLC ("Netgate")
Differential Revision: https://reviews.freebsd.org/D38746
Fixes the following warning:
yacc: w - the symbol System_spec is undefined
yacc: 3 rules never reduced
Reported by: otis
Fixes: 6a836ea741 ("config(8): Remove obsolete 'config' directive.")
For compatibility with glibc. The previous code would trigger a division
by zero in roundup() and terminate. Instead, just pass through to
malloc() for align == 0.
PR: 269688
Reviewed by: imp, mjg
MFC after: 1 week
Pull Request: https://github.com/freebsd/freebsd-src/pull/655
This file defines a small API to be used by other drivers. If any of
these functions are called before the bcm_dma device has attached we
should handle the error gracefully. Fix a formatting quirk while here.
Reviewed by: manu
MFC after: 1 week
Differential Revision: https://reviews.freebsd.org/D38756
The sdhci_bcm driver attach routine relies on bcm_dma already being
attached, in order to allocate a DMA channel. However, both drivers
attached at the default pass so this is not guaranteed. Newer RPI
firmware exposes this assumption, and the result is a NULL-dereference
in bcm_dma_allocate().
To fix this, use BUS_PASS_SUPPORTDEV for bcm_dma.
PR: 268835
Reviewed by: mhorne
MFC after: 1 week
The Kasan tests show the nfsrvd_cleancache() results
in a modify after free. I think this occurs because the
nfsrv_cleanup() function gets executed after nfs_cleanup()
which free's the nfsstatsv1_p.
This patch makes them use the same subsystem and sets
SI_ORDER_FIRST for nfs_cleanup(), so that it will be called
after nfsrv_cleanup() via VNET_SYSUNINIT().
The patch also sets nfsstatsv1_p NULL after free'ng it,
so that a crash will result if it is used after free'ng.
Tested by: markj
Reviewed by: markj
MFC after: 3 months
Differential Revision: https://reviews.freebsd.org/D38750
ping -R (F_RROUTE) will loop at ping.c:1381 until it segfaults or
the unsigned int hlen happens to be less than the size of an IP header:
slippy$ ping -R 192.168.0.101
PING 192.168.0.101 (192.168.0.101): 56 data bytes
64 bytes from 192.168.0.101: icmp_seq=0 ttl=63 time=1.081 ms
RR: 192.168.0.1
192.168.0.101
192.168.0.101
10.1.1.254
10.1.1.91
unknown option bb
unknown option 32
unknown option 6
...
unknown option 96
unknown option 2d
Segmentation fault
The reason for this is while looping through loose source routing (LSRR)
and strict source routing (SSRR), hlen will become smaller than the IP
header. It may even become negative. This should terminate the loop.
However, when hlen is unsigned, an integer underflow occurs becoming a
large number causing the loop to continue virtually forever until hlen
is either by chance smaller than the lenghth of an IP header or it
segfaults.
Reviewed by: asomers
Fixes: 46d7b45a26
MFC after: 3 days
Differential Revision: https://reviews.freebsd.org/D38744
Most options in kernel config files use "options<space><tab>OPTION".
This allows the option to be commented out without shifting columns.
A few options had two tabs, and some had spaces. Make them consistent.
Most options in kernel config files use "options<space><tab>OPTION".
This allows the option to be commented out without shifting columns.
A few options had two tabs, and some had spaces. Make them consistent.
Most options in kernel config files use "options<space><tab>OPTION".
This allows the option to be commented out without shifting columns.
A few options had two tabs, and some had spaces. Make them consistent.
Most options in kernel config files use "options<space><tab>OPTION".
This allows the option to be commented out without shifting columns.
A few options had two tabs, and some had spaces. Make them consistent.
Most options in kernel config files use "options<space><tab>OPTION".
This allows the option to be commented out without shifting columns.
A few options had two tabs, and some had spaces. Make them consistent.
