condition was addressed in libpuffs by counting lookups.
The fix assumes that cookies map to struct puffs_cookie, which has not
been documented as a requirement for filesystems using libpuffs. As an
example, we got burnt by this assumption in libp2k (kern/46734), and
we fixed bit by actually mapping libp2k cookies to struct puffs_node.
It is unlikely, but there may be third party filesystems that use cookies
unmapped to struct puffs_node, and they were left broken for now.
- we introduce a puffs_init() flag PUFFS_FLAG_PNCOOKIE that let filesystems
inform libpuffs that they map cookies to struct puffs_node. Is that flag
is used, the lookup/reclaim race condition fix is enabled. We enable the
flag for libp2k.
- filesystems that use puffs_pn_new() obviouslty use struct puffs_node
and gain PUFFS_FLAG_PNCOOKIE automatically even if they did not specify
it in puffs_init(). This include all our PUFFS filesystem in-tree except
libp2k.
- for filesystems not willing to use struct puffs_node, we introduce a
reclaim2 vnop, which is reclaim with an additionnal lookup count argument.
This vnop let the filesystem implement the lookup/reclaim race fix on
its own.
pthread__smt_pause. These are implemented using the ARM instructions
SEV (wake) and WFE (pause). These are treated as NOPs on ARM CPUs that
don't support them.
1 before return when invoked from m_xword() and m_xhalf() functions.
The caller doesn't set it to 1 either. So, set *err to 1 before
invoking MINDEX().
- When handling contiguous buffer in _bpf_tap(), pass its real size
rather than 0 to avoid reading packet data as mbuf struct on
out-of-bounds loads.
- Correctly pass pktlen and buflen arguments from bpf_deliver() to
bpf_filter() to avoid reading mbuf struct as packet data.
JIT case is still broken.
Also, test pointers againts NULL.
as sys/dev/sun/event_var.h does.
Fixes immediate KASSERT(mutex_owned(proc_lock)) panic during running Xserver
on DIAGNOSTIC kernel.
Should be pulled up to netbsd-6.
- get rid of some redundant variables ( sc_*phys and sc_*base were essentially
the same )
- we don't access video memory directly anymore, so:
* no more need for byte order tests
* don't map video memory if we have a MMIO register aperture
* get rid of various bits of support code dealing with video memory
access
- use MMIO register aperture if available, turn off register blocks in the main
aperture if not in use, to avoid overlap with video memory on 8MB cards
- don't scissor off access to off-screen memory. How the hell did this work on
Rage II? Apparently Rage XL treats the scissor registers slightly different.
Now this works properly on 8MB and Rage XL boards like Sun's PGX64.
(this is not entirely correct because it can return ENOMEM which is
not mentioned in the spec, but there are other places in pthread_create
whete ENOMEM is returned -- it at all, this should be fixed everywhere)
That uses devname(3) internally, which doesn't work at all in a cross
build environment, and doesn't do what I thought even in a native
environment.
Instead, parse the device major numbers for the pty master and slave
devices from the output of "MAKEDEV -s pty0" and check those against the
actual device node that we are thinking of removing.
- downgrade the error of not finding a file to load to a debug message.
Statically linked OpenPAM (like on sun2) does not have shared objects.
- make sure we preserve errno around dlclose() which will call munmap()
XXX: Pullup to 6
