compiler optimising away or reordering accesses to them.
Fixes ehci(4) on NetBSD/evbarm when using -Os optimisation. It's likely
ohci(4) and uhci(4) are similarly afflicted, so the same changes have
been made there.
Quite how other platforms got away without this for so long is a mystery...
wrappers around the speicificdata subroutines. Also:
- Call the new lwpinit() function from main() after calling procinit().
- Move some pool initialization out of kern_proc.c and into files that
are directly related to the pools in question (kern_lwp.c and kern_ras.c).
- Convert uipc_sem.c to proc_{get,set}specific(), and eliminate the p_ksems
member from struct proc.
subsystems, based on work by YAMAMOTO Takashi. This is intended to
be used by other subsystems (such as the proc_*() or lwp_*() routines)
rather than directly by consumers.
chip
We did some digging, and the Netgear WG311 has three versions. The one
supported by this driver does, in fact, have an Atheros chip. V3 was
already listed as not supported, and the card you are referring to is
a V2 (with the TI chip) -- so a note was added right above the note for
the V3 version.
Input from David A. Holland and mlelstv@ on #NetBSD-code, thanks!
Apply OpenBSD src/sys/net/pf.c rev 1.486 and 1.487:
1.486:
When synproxy sends packets to the destination host, make sure to copy
the 'tag' from the original state entry into the outgoing mbuf.
1.487:
When synproxy completes the replayed handshake and modifies the state
into a normal one, it sets both peers' sequence windows. Fix a bug where
the previously advertised windows are applied to the wrong side (i.e.
peer A's seqhi is peer A's seqlo plus peer B's, not A's, window). This
went undetected because mostly the windows are similar and/or re-
advertised soon. But there are (rare) cases where a synproxy'd connection
would stall right after handshake. Found by Gleb Smirnoff.
blitting. Thanks to David Redman (Tadpole) for noticing it. This probably
escaped notice before, since we never do overlapping blits (in the X
direction), but this fix may prevent problems if someone ever does use it
for that.
We had a (long) discussion about this in #NetBSD-code and the consensus was
that /tmp should always be sticky (+t), and as such it is safe to create
the temporary files for chpass(1) in there. We added a check to guarantee
this (and bail out otherwise) and now temporary files are created in /tmp,
not filling up /etc.
Thanks to rivo nurges et al.
- remove any 'permission checks' via geteuid() - with upcoming
security models these might not match the model any more
- this also fixes a bug where ntp_adjtime() was denied (EPERM)
even though only a legit read status was performed
all callers of these functions are at splsoftnet already:
tcp_sack_option
tcp_input ok
tcp_del_sackholes
tcp_input ok
tcp_free_sackholes
tcp_close ok
tcp_timer_rexmt ok
tcp_timer_2msl ok
all callers of tcp_close are at splsoftnet already:
tcp_close
tcp_input ok
tcp_disconnect
tcp_usrreq ok
tcp_usrclosed
tcp_usrreq ok
tcp_disconnect
tcp_timer_2msl ok
tcp_drop
tcp_usrreq
tcp_disconnect
tcp_timer_rexmt ok
tcp_timer_persist ok
tcp_timer_keep ok
tcp_input
syn_cache_get
tcp_input
