Skip to content
Permalink
Tree: e2502dba36
Commits on Apr 20, 2019
  1. The rip6 checksum errors are accounted per socket. So the no socket

    bluhm committed Apr 20, 2019
    errors include these, do not subtract both from delivered.  Avoids
    an underflow in the rip6 delivered counter.
    OK deraadt@ claudio@
  2. Statistics of "netstat -s -f inet6 -p rip6" did not work. In

    bluhm committed Apr 20, 2019
    rip6_sysctl_rip6stat() copy out rip6counters, not ip6counters.
    OK deraadt@ claudio@
  3. Restore previous section after setting the MIPS ABI marker. This keeps

    visa
    visa committed Apr 20, 2019
    the .text section in use after the file header, improving compatibility
    with gcc. Without this change, module-level inline assembly blocks could
    end up into wrong section.
    
    OK kettenis@ guenther@
  4. Avoid undefined behaviour that results from negating a signed long with

    4a6f656c committed Apr 20, 2019
    minimum value.
    
    Fixes oss-fuzz #14354.
    
    ok beck@ bcook@ tb@
  5. drm/ttm: Fix bo_global and mem_global kfree error

    jonathangray committed Apr 20, 2019
    From Trigger Huang
    b10cc08374728ea79555a1cd98f962b0f942e942 in linux 4.19.y/4.19.36
    30f33126feca0fe16df9e9302ffc28a953e2eb37 in mainline linux
  6. Move lockf structures from header to implementation since external users

    anton
    anton committed Apr 20, 2019
    only need a lockf_state pointer by now.
    
    ok mpi@ visa@
  7. Adapt radeondrm_detach_kms() to struct drm_device being split from

    jonathangray committed Apr 20, 2019
    drm softc.
    
    Avoids uvm_fault() when firmware is missing and radeondrm is forcibly
    detached.  Reported by Mihai Popescu on misc@
Commits on Apr 19, 2019
  1. On powerpc, link with -Wl,-relax as clang is a monster and code excee…

    kettenis committed Apr 19, 2019
    …ds the
    
    maximum reachability of the PowerPC branch instructions.
    
    Also override NOPIE_FLAGS to avoid building code with -fno-pie as doing so
    is incompatible with secure-plt when using clang as the compiler.
    
    ok visa@, guenther@
  2. Start a realpath(2) regress test, currently not enabled.

    beck
    beck committed Apr 19, 2019
    This will extend as I add more cases to it. It will come
    into use once an in-kernel version of realpath(2) becomes
    a real thing
  3. Allocate md_data with calloc to avoid use of uninitialised memory.

    4a6f656c committed Apr 19, 2019
    Found by Guido Vranken when fuzzing and trying to use GOST with HMAC.
    
    Fix confirmed by Guido; ok tb@
  4. Add a subsystem lock for vfs_lockf.c. This enables calling lf_advlock()

    visa
    visa committed Apr 19, 2019
    and lf_purgelocks() without the kernel lock.
    
    OK anton@ mpi@
  5. Work around a limitation of clang integrated assembler on mips64.

    visa
    visa committed Apr 19, 2019
    The assembler does not handle undeclared local symbols properly
    and generates R_MIPS_CALL16 relocations where it should generate
    local GOT references. For now, get along with the problem by
    declaring local symbols where necessary.
    
    OK kettenis@ guenther@
  6. Prevent clang from using builtins and jump tables in _dl_boot_bind()

    visa
    visa committed Apr 19, 2019
    on mips64. They need relocation and consequently cannot be used
    in that function.
    
    OK kettenis@
  7. use the common code in if.c to check if txhprio is good.

    dlg
    dlg committed Apr 19, 2019
    no functional change.
  8. use the factored out txhprio and rxhprio checks

    dlg
    dlg committed Apr 19, 2019
    reduces code duplication and chance for error.
  9. provide factored out txhprio and rxhprio checks

    dlg
    dlg committed Apr 19, 2019
    l2 and l3 drivers do the same thing all the time, so reduce the
    chance of error by doing the checks once and making it available
    for drivers to call instead of rolling on their own again.
  10. add rxprio support

    dlg
    dlg committed Apr 19, 2019
    this is modelled on vlan(4) where the packet prio is put in the bpe
    header in tx, and the bpe header prio is put on the packet in rx.
  11. add rxprio support

    dlg
    dlg committed Apr 19, 2019
  12. implement rxprio

    dlg
    dlg committed Apr 19, 2019
  13. allow configuration of rxprio

    dlg
    dlg committed Apr 19, 2019
    ok claudio@
  14. add support for configuring rxprio.

    dlg
    dlg committed Apr 19, 2019
    vlan already used the 802.1p prio in packets to set the mbuf prio.
    this maintains that as the default.
    
    ok claudio@
  15. rxprio.

    dlg
    dlg committed Apr 19, 2019
  16. add support for getting and setting rxprio

    dlg
    dlg committed Apr 19, 2019
    this complements txprio and should finish support for RFC 2983
    
    ok claudio@
  17. only root can change rxprio

    dlg
    dlg committed Apr 19, 2019
  18. add SIOCSRXHPRIO and SIOCGRXHPRIO for configuring rx prio handling

    dlg
    dlg committed Apr 19, 2019
    this is the complement of txprio handling, and helps support RFC 2983.
    
    ok claudio@
  19. add IF_HDRPRIO_OUTER for rxprio

    dlg
    dlg committed Apr 19, 2019
    IF_HDRPRIO_OUTER says you want the priority from the outer encap header.
    
    ok claudio@
Commits on Apr 18, 2019
  1. Always check for namespace collisions on table commands

    kn
    kn committed Apr 18, 2019
    `-t table -T add|replace ...' would only check for duplicate tables in case
    addresses where actually to the table.
    
    Instead of using a positive number of added addresses as prove for
    successful table operations, rely on the fact that CREATE_TABLE() is
    guaranteed to be called only if pf(4) can be accessed, that is
    warn_duplicate_tables() will return.
    
    This improves duplicate detection rate as warnings are now also emitted
    even when table commands eventually leave tables unchanged.
    
    OK benno sashan
  2. Fix table definition parsing as unprivileged user

    kn
    kn committed Apr 18, 2019
    revision 1.689 introduced warn_duplicate_tables() unconditionally, breaking
    the parser on tables withs insufficient permissions to open pf(4):
    
    	$ echo 'table <t>' | pfctl -nf-
    	pfctl: pfr_get_tables: Bad file descriptor
    
    So simply check whether pfctl is able to get the table list first.  If not,
    instead of silently avoiding namespace collision checks, print a brief
    notice iff `-v' is given to help finding duplicate definitions by hand:
    
    	$ echo 'table <t>' | ./obj/pfctl -vnf-
    	table <t>
    	stdin:1: skipping duplicate table checks for <t>
    
    Reported by Rivo Nurges, thanks!
    OK benno sashan
Older
You can’t perform that action at this time.