Merge branch 'freebsd/14-stable/main' into hardened/14-stable/master
| Delta | File | |
|---|---|---|
| +19 | -4 | sys/dev/ixgbe/ixgbe_phy.c |
| +4 | -1 | sys/dev/ixgbe/if_ix.c |
| +3 | -1 | sys/dev/ixgbe/ixgbe_82599.c |
| +1 | -2 | sys/dev/e1000/em_txrx.c |
| +2 | -0 | sys/dev/ixgbe/ixgbe_phy.h |
| +2 | -0 | sys/dev/ixgbe/ixgbe_type.h |
| +31 | -8 | 2 files not shown |
| +31 | -10 | 8 files |
e1000: Fix vlan PCP/DEI on lem(4) The vlan PCP and CFI/DEI were discarded when receiving vlan tagged packets on lem(4) interfaces with vlanhwtag. According to the 82540 SDM[1] (pg. 24), vlan tag is in the standard format, so there's no reason to discard PCP/DEI. [1]: http://iommu.com/datasheets/ethernet/controllers-nics/intel/e1000/pci-pci-x-family-gbe-controllers-software-dev-manual.pdf Sponsored by: Stormshield (author) Differential Revision: https://reviews.freebsd.org/D48987 (cherry picked from commit 4b29599fbbe33b75b7b58cfc5deea7a881e9a10e)
| Delta | File | |
|---|---|---|
| +1 | -2 | sys/dev/e1000/em_txrx.c |
| +0 | -1 | sys/dev/e1000/e1000_defines.h |
| +1 | -3 | 2 files |
Merge remote-tracking branch 'origin/hardened/current/master' into hardened/current/cross-dso-cfi
| Delta | File | |
|---|---|---|
| +118 | -98 | sys/netpfil/pf/pf.c |
| +151 | -4 | tests/sys/netpfil/pf/src_track.sh |
| +26 | -14 | sys/netpfil/pf/pf_ioctl.c |
| +20 | -15 | sys/netpfil/pf/pf_lb.c |
| +23 | -7 | sys/net/pfvar.h |
| +22 | -3 | sys/netpfil/pf/pf_nl.c |
| +360 | -141 | 13 files not shown |
| +411 | -166 | 19 files |
Merge branch 'freebsd/current/main' into hardened/current/master
| Delta | File | |
|---|---|---|
| +118 | -98 | sys/netpfil/pf/pf.c |
| +151 | -4 | tests/sys/netpfil/pf/src_track.sh |
| +26 | -14 | sys/netpfil/pf/pf_ioctl.c |
| +20 | -15 | sys/netpfil/pf/pf_lb.c |
| +23 | -7 | sys/net/pfvar.h |
| +22 | -3 | sys/netpfil/pf/pf_nl.c |
| +360 | -141 | 13 files not shown |
| +411 | -166 | 19 files |
e1000: Fix vlan PCP/DEI on lem(4) The vlan PCP and CFI/DEI were discarded when receiving vlan tagged packets on lem(4) interfaces with vlanhwtag. According to the 82540 SDM[1] (pg. 24), vlan tag is in the standard format, so there's no reason to discard PCP/DEI. [1]: http://iommu.com/datasheets/ethernet/controllers-nics/intel/e1000/pci-pci-x-family-gbe-controllers-software-dev-manual.pdf MFC after: 3 days Sponsored by: Stormshield (author) Differential Revision: https://reviews.freebsd.org/D48987
| Delta | File | |
|---|---|---|
| +1 | -2 | sys/dev/e1000/em_txrx.c |
| +0 | -1 | sys/dev/e1000/e1000_defines.h |
| +1 | -3 | 2 files |
Merge branch 'freebsd/14-stable/main' into hardened/14-stable/master
| Delta | File | |
|---|---|---|
| +620 | -426 | sys/dev/e1000/if_em.c |
| +469 | -345 | sys/dev/ixgbe/if_ix.c |
| +247 | -204 | sys/dev/igc/if_igc.c |
| +137 | -117 | sys/dev/ixgbe/if_ixv.c |
| +56 | -54 | sys/dev/ixgbe/if_bypass.c |
| +57 | -49 | sys/dev/ixgbe/ix_txrx.c |
| +1,586 | -1,195 | 6 files not shown |
| +1,719 | -1,297 | 12 files |
e1000: Style txrx Fix up indentation and reflow long lines. Sponsored by: BBOX.io (cherry picked from commit 7390daf87c53073da5e74b2d73ac7320c1a44475)
| Delta | File | |
|---|---|---|
| +34 | -21 | sys/dev/e1000/em_txrx.c |
| +26 | -18 | sys/dev/e1000/igb_txrx.c |
| +2 | -1 | sys/dev/e1000/if_em.c |
| +62 | -40 | 3 files |
Merge remote-tracking branch 'origin/hardened/current/master' into hardened/current/cross-dso-cfi
| Delta | File | |
|---|---|---|
| +474 | -664 | sys/contrib/openzfs/module/zfs/brt.c |
| +469 | -345 | sys/dev/ixgbe/if_ix.c |
| +137 | -117 | sys/dev/ixgbe/if_ixv.c |
| +56 | -54 | sys/dev/ixgbe/if_bypass.c |
| +50 | -58 | sys/contrib/openzfs/include/sys/brt_impl.h |
| +63 | -45 | sys/contrib/openzfs/module/zfs/zap_leaf.c |
| +1,249 | -1,283 | 43 files not shown |
| +1,818 | -1,583 | 49 files |
Merge branch 'freebsd/current/main' into hardened/current/master
| Delta | File | |
|---|---|---|
| +474 | -664 | sys/contrib/openzfs/module/zfs/brt.c |
| +469 | -345 | sys/dev/ixgbe/if_ix.c |
| +137 | -117 | sys/dev/ixgbe/if_ixv.c |
| +56 | -54 | sys/dev/ixgbe/if_bypass.c |
| +63 | -45 | sys/contrib/openzfs/module/zfs/zap_leaf.c |
| +50 | -58 | sys/contrib/openzfs/include/sys/brt_impl.h |
| +1,249 | -1,283 | 43 files not shown |
| +1,818 | -1,583 | 49 files |
e1000: Style txrx Fix up indentation and reflow long lines. MFC after: 3 days Sponsored by: BBOX.io
| Delta | File | |
|---|---|---|
| +34 | -21 | sys/dev/e1000/em_txrx.c |
| +26 | -18 | sys/dev/e1000/igb_txrx.c |
| +2 | -1 | sys/dev/e1000/if_em.c |
| +62 | -40 | 3 files |
Merge branch 'freebsd/14-stable/main' into hardened/14-stable/master
| Delta | File | |
|---|---|---|
| +579 | -504 | sys/dev/igc/if_igc.c |
| +758 | -29 | sys/arm64/arm64/vfp.c |
| +250 | -0 | sys/arm64/include/armreg.h |
| +135 | -0 | sys/dev/random/armv8rng.c |
| +83 | -15 | sys/arm64/arm64/identcpu.c |
| +88 | -2 | sys/arm64/arm64/exec_machdep.c |
| +1,893 | -550 | 31 files not shown |
| +2,181 | -721 | 37 files |
Merge branch 'freebsd/13-stable/main' into hardened/13-stable/master
| Delta | File | |
|---|---|---|
| +579 | -504 | sys/dev/igc/if_igc.c |
| +35 | -34 | usr.sbin/freebsd-update/freebsd-update.8 |
| +33 | -35 | sys/dev/igc/igc_txrx.c |
| +16 | -19 | sys/dev/e1000/em_txrx.c |
| +14 | -16 | sys/dev/e1000/igb_txrx.c |
| +14 | -14 | sys/dev/igc/if_igc.h |
| +691 | -622 | 2 files not shown |
| +692 | -629 | 8 files |
e1000: txrx function prototype cleanup Drop variable names of function prototypes since the file is mixed in listing them or not and they fall out of sync. Sponsored by: BBOX.io (cherry picked from commit 9dc452b983f042aa9d2a2562517d44c1928ff42a)
| Delta | File | |
|---|---|---|
| +16 | -19 | sys/dev/e1000/em_txrx.c |
| +14 | -16 | sys/dev/e1000/igb_txrx.c |
| +30 | -35 | 2 files |
e1000: txrx function prototype cleanup Drop variable names of function prototypes since the file is mixed in listing them or not and they fall out of sync. Sponsored by: BBOX.io (cherry picked from commit 9dc452b983f042aa9d2a2562517d44c1928ff42a)
| Delta | File | |
|---|---|---|
| +16 | -19 | sys/dev/e1000/em_txrx.c |
| +14 | -16 | sys/dev/e1000/igb_txrx.c |
| +30 | -35 | 2 files |
Merge branch 'freebsd/14-stable/main' into hardened/14-stable/master
| Delta | File | |
|---|---|---|
| +264 | -5 | sys/dev/e1000/if_em.c |
| +21 | -7 | sys/dev/e1000/if_em.h |
| +10 | -0 | sys/dev/e1000/em_txrx.c |
| +2 | -2 | sys/conf/newvers.sh |
| +4 | -0 | sys/dev/e1000/igb_txrx.c |
| +1 | -1 | usr.bin/yacc/config.h |
| +302 | -15 | 6 files |
Merge branch 'freebsd/13-stable/main' into hardened/13-stable/master
| Delta | File | |
|---|---|---|
| +264 | -5 | sys/dev/e1000/if_em.c |
| +21 | -7 | sys/dev/e1000/if_em.h |
| +10 | -0 | sys/dev/e1000/em_txrx.c |
| +4 | -0 | sys/dev/e1000/igb_txrx.c |
| +299 | -12 | 4 files |
e1000: Re-add AIM
We originally left this out because iflib modulates interrupts and
accomplishes some level of batching versus the custom queues in the
older driver. Upon more detailed study of the Linux driver which has a
newer implementation, it finally became clear to me this is actually a
holdoff timer and not an interrupt limit as it is conventionally
(statically) programmed and displayed as an interrupt rate. The data
sheets also make this somewhat clear.
Thus, AIM accomplishes two beneficial things for a wide variety of
workloads[1]:
1. At low throughput/packet rates, it will significantly lower latency
(by counter-intuitively "increasing" the interrupt rate.. better
thought of as decreasing the holdoff timer because you will modulate
down before coming anywhere near these interrupt rates).
2. At bulk data rates, it is tuned to achieve a lower interrupt rate
(by increasing the holdoff timer) than the current static 8000/s. This
[23 lines not shown]| Delta | File | |
|---|---|---|
| +264 | -5 | sys/dev/e1000/if_em.c |
| +21 | -7 | sys/dev/e1000/if_em.h |
| +10 | -0 | sys/dev/e1000/em_txrx.c |
| +4 | -0 | sys/dev/e1000/igb_txrx.c |
| +299 | -12 | 4 files |
e1000: Re-add AIM
We originally left this out because iflib modulates interrupts and
accomplishes some level of batching versus the custom queues in the
older driver. Upon more detailed study of the Linux driver which has a
newer implementation, it finally became clear to me this is actually a
holdoff timer and not an interrupt limit as it is conventionally
(statically) programmed and displayed as an interrupt rate. The data
sheets also make this somewhat clear.
Thus, AIM accomplishes two beneficial things for a wide variety of
workloads[1]:
1. At low throughput/packet rates, it will significantly lower latency
(by counter-intuitively "increasing" the interrupt rate.. better
thought of as decreasing the holdoff timer because you will modulate
down before coming anywhere near these interrupt rates).
2. At bulk data rates, it is tuned to achieve a lower interrupt rate
(by increasing the holdoff timer) than the current static 8000/s. This
[23 lines not shown]| Delta | File | |
|---|---|---|
| +264 | -5 | sys/dev/e1000/if_em.c |
| +21 | -7 | sys/dev/e1000/if_em.h |
| +10 | -0 | sys/dev/e1000/em_txrx.c |
| +4 | -0 | sys/dev/e1000/igb_txrx.c |
| +299 | -12 | 4 files |
Merge remote-tracking branch 'origin/hardened/current/master' into hardened/current/cross-dso-cfi
| Delta | File | |
|---|---|---|
| +579 | -504 | sys/dev/igc/if_igc.c |
| +33 | -35 | sys/dev/igc/igc_txrx.c |
| +36 | -18 | sys/dev/netmap/netmap_mem2.c |
| +18 | -32 | sys/kern/vfs_subr.c |
| +46 | -0 | sys/compat/linuxkpi/common/include/linux/cleanup.h |
| +16 | -19 | sys/dev/e1000/em_txrx.c |
| +728 | -608 | 144 files not shown |
| +1,728 | -765 | 150 files |
Merge branch 'freebsd/current/main' into hardened/current/master
| Delta | File | |
|---|---|---|
| +579 | -504 | sys/dev/igc/if_igc.c |
| +33 | -35 | sys/dev/igc/igc_txrx.c |
| +36 | -18 | sys/dev/netmap/netmap_mem2.c |
| +18 | -32 | sys/kern/vfs_subr.c |
| +46 | -0 | sys/compat/linuxkpi/common/include/linux/cleanup.h |
| +16 | -19 | sys/dev/e1000/em_txrx.c |
| +728 | -608 | 144 files not shown |
| +1,728 | -765 | 150 files |
e1000: txrx function prototype cleanup Drop variable names of function prototypes since the file is mixed in listing them or not and they fall out of sync. MFC after: 1 week Sponsored by: BBOX.io
| Delta | File | |
|---|---|---|
| +16 | -19 | sys/dev/e1000/em_txrx.c |
| +14 | -16 | sys/dev/e1000/igb_txrx.c |
| +30 | -35 | 2 files |
Merge remote-tracking branch 'origin/hardened/current/master' into hardened/current/cross-dso-cfi
| Delta | File | |
|---|---|---|
| +264 | -5 | sys/dev/e1000/if_em.c |
| +21 | -7 | sys/dev/e1000/if_em.h |
| +10 | -0 | sys/dev/e1000/em_txrx.c |
| +4 | -0 | sys/dev/e1000/igb_txrx.c |
| +299 | -12 | 4 files |
Merge branch 'freebsd/current/main' into hardened/current/master
| Delta | File | |
|---|---|---|
| +264 | -5 | sys/dev/e1000/if_em.c |
| +21 | -7 | sys/dev/e1000/if_em.h |
| +10 | -0 | sys/dev/e1000/em_txrx.c |
| +4 | -0 | sys/dev/e1000/igb_txrx.c |
| +299 | -12 | 4 files |
e1000: Re-add AIM
We originally left this out because iflib modulates interrupts and
accomplishes some level of batching versus the custom queues in the
older driver. Upon more detailed study of the Linux driver which has a
newer implementation, it finally became clear to me this is actually a
holdoff timer and not an interrupt limit as it is conventionally
(statically) programmed and displayed as an interrupt rate. The data
sheets also make this somewhat clear.
Thus, AIM accomplishes two beneficial things for a wide variety of
workloads[1]:
1. At low throughput/packet rates, it will significantly lower latency
(by counter-intuitively "increasing" the interrupt rate.. better
thought of as decreasing the holdoff timer because you will modulate
down before coming anywhere near these interrupt rates).
2. At bulk data rates, it is tuned to achieve a lower interrupt rate
(by increasing the holdoff timer) than the current static 8000/s. This
[22 lines not shown]| Delta | File | |
|---|---|---|
| +264 | -5 | sys/dev/e1000/if_em.c |
| +21 | -7 | sys/dev/e1000/if_em.h |
| +10 | -0 | sys/dev/e1000/em_txrx.c |
| +4 | -0 | sys/dev/e1000/igb_txrx.c |
| +299 | -12 | 4 files |
Merge remote-tracking branch 'internal/freebsd/13-stable/main' into hardened/13-stable/master
Conflicts:
bin/cat/cat.c
lib/libc/gen/auxv.3
libexec/rc/rc.d/archdep
share/man/man5/src.conf.5
sys/amd64/amd64/elf_machdep.c
sys/amd64/cloudabi64/cloudabi64_sysvec.c
sys/amd64/linux/linux_sysvec.c
sys/amd64/linux32/linux32_sysvec.c
sys/arm/arm/elf_machdep.c
sys/arm64/linux/linux_sysvec.c
sys/compat/ia32/ia32_sysvec.c
sys/i386/cloudabi32/cloudabi32_sysvec.c
sys/i386/linux/imgact_linux.c
sys/i386/linux/linux_sysvec.c
sys/kern/imgact_aout.c
sys/kern/kern_module.c
[22 lines not shown]| Delta | File | |
|---|---|---|
| +175 | -44 | sys/vm/vm_map.c |
| +97 | -38 | sys/kern/vfs_subr.c |
| +64 | -0 | tests/sys/vm/stack/stack_dlopen_exec_test.c |
| +57 | -0 | tests/sys/vm/stack/stack_mprotect_exec_test.c |
| +42 | -11 | contrib/less/screen.c |
| +51 | -0 | tests/sys/vm/mmap_map_32bit_helper.c |
| +486 | -93 | 31,889 files not shown |
| +1,170 | -49,853 | 31,895 files |
sys: Remove $FreeBSD$: one-line .c comment pattern Remove /^/[*/]\s*\$FreeBSD\$.*\n/ Similar commit in current: (cherry picked from commit 71625ec9ad2a)
Merge branch 'freebsd/13-stable/main' into hardened/13-stable/master
| Delta | File | |
|---|---|---|
| +40 | -29 | sys/dev/e1000/if_em.c |
| +3 | -0 | sys/kern/kern_osd.c |
| +1 | -1 | sys/net/if_vlan.c |
| +1 | -1 | sys/dev/e1000/em_txrx.c |
| +45 | -31 | 4 files |
e1000: Fix off by one ipcse This has been off by one in the FreeBSD drivers as far back as I've looked. Emperically HW and SW emulations I have available don't seem to mind. Noticed while debugging other issues. (cherry picked from commit eac761e967b34066aa7183474249df87d79a2f32)
| Delta | File | |
|---|---|---|
| +1 | -1 | sys/dev/e1000/em_txrx.c |
| +1 | -1 | 1 files |
Merge remote-tracking branch 'internal/hardened/current/master' into hardened/current/cross-dso-cfi
Conflicts:
tools/build/options/WITH_OFED_EXTRA (deleted)
usr.sbin/rpc.yppasswdd/Makefile (unresolved)
| Delta | File | |
|---|---|---|
| +697 | -0 | sys/dev/virtio/gpu/virtio_gpu.c |
| +454 | -0 | sys/dev/virtio/gpu/virtio_gpu.h |
| +153 | -0 | sys/arm64/arm64/pl031_rtc.c |
| +80 | -52 | sys/compat/linux/linux_socket.c |
| +107 | -0 | tools/build/freebsd-yeet.pl |
| +87 | -11 | sys/dev/ixl/ixl_pf_main.c |
| +1,578 | -63 | 31,042 files not shown |
| +2,188 | -48,405 | 31,048 files |
Merge remote-tracking branch 'internal/freebsd/current/main' into hardened/current/master
Conflicts:
bin/cat/cat.c (unresolved)
lib/csu/aarch64/crt1_c.c (unresolved)
lib/csu/amd64/crt1_c.c (unresolved)
lib/csu/i386/crt1_c.c (unresolved)
lib/csu/powerpc64/crt1_c.c (unresolved)
lib/csu/powerpc64/reloc.c (unresolved)
lib/csu/riscv/crt1_c.c (unresolved)
lib/libc/gen/auxv.3 (unresolved)
lib/libcasper/services/cap_net/cap_net.3 (unresolved)
sys/amd64/amd64/elf_machdep.c (unresolved)
sys/amd64/include/sgxreg.h (deleted)
sys/amd64/linux/linux_sysvec.c (unresolved)
sys/amd64/linux32/linux32_sysvec.c (unresolved)
sys/amd64/sgx/sgx.c (deleted)
sys/amd64/sgx/sgx_linux.c (deleted)
sys/amd64/sgx/sgxvar.h (deleted)
[29 lines not shown]| Delta | File | |
|---|---|---|
| +697 | -0 | sys/dev/virtio/gpu/virtio_gpu.c |
| +454 | -0 | sys/dev/virtio/gpu/virtio_gpu.h |
| +153 | -0 | sys/arm64/arm64/pl031_rtc.c |
| +80 | -52 | sys/compat/linux/linux_socket.c |
| +107 | -0 | tools/build/freebsd-yeet.pl |
| +87 | -11 | sys/dev/ixl/ixl_pf_main.c |
| +1,578 | -63 | 31,044 files not shown |
| +2,304 | -48,375 | 31,050 files |