cc: get rid of sprint() and temporary buffer for type conversion fuction
slookup() copies to symb, so use the symb[NSYMB] buffer directly
to declare type conversion functions and get rid of the arbitrary
sized local buffer. replace sprint() with snprint().
6c: conserve registers for floating point operations
for floating point operations, reuse the return register
on the right hand side if it has higher complex number
than the left hand side to conserve registers.
this makes the following code compile, that was previously
run out of floating point register:
rc: fix code serialization for PIPEFD (thanks BurnZeZ)
BurnZeZ reported this the other day. It seems like if we have
a pipeline that looks like:
fn foo{cat < <{echo hi}}
then the '<' will get merged in /env/'fn#foo'. This change
fixes pcmd to add a space. It looks to me like this is the
only token that can get merged this way by pcmd.
Ori Bernstein [Sat, 11 Apr 2020 21:19:46 +0000 (14:19 -0700)]
triple-click to select non-whitespace segment
The previous patch to plumb non-whitespace segments was
confusing due to lack of visual feedback. This removes
the empty selecton plumb behavior, and instead makes
triple clicking work to get a plumbable selection.
when running ndb configuration, we might inherit the ipgw=
attribute from the ipnet pointing to our own ip address
(we are the default gateway). ignore such entries.
do not add default routes with gateway equal to our own
local (ip4) or link-local ip address (ipv6).
cc, ?[acl]: fix gethunk() and move common memory allocator code to cc/compat
for gethunk() to work, all allocators have to use it,
including allocations done by libc thru malloc(),
so the fake allocation functions are mandatory for
everyone.
to avoid duplication the code is moved to cc/compat
and prototypes provided in new cc/compat.h header.
backout the gethunk() again, as that breaks the assemblers
the assemblers share gethunk() cc/macbody but are compiled
without compat.c, so calls such as getenv() trigger malloc()
which does its own sbrk() calls, breaking the continuity
of the hunk.
so this change needs another revision. until then, this is
backed out.
the compilers and linkers use ther own memory allocator.
free memory is between hunk and hunk+nhunk. allocation
works by checking if nhunk is bigger or equal to the
amount needed, and if not, repeatedly call gethunk()
until there is. after that, the allocated amount is added
from hunk and subtracted from nhunk by the user.
the problem was when the needed amount was bigger than
the default NHUNK size gethunk() allocates per call.
gethunk() would not actually grow nhunk, but instead
just set hunk and nhunk variables to the last allocated
block. this resulted in a infinite loop of calls to
gethunk() until sbrk() would hit the maximum size for
the BSS segment.
this change makes gethunk() actually grow the hunk space,
increasing nhunk, and only updating hunk when nhunk was
previously zero. we assume that mysbrk() retuns increasing
addresses and that the space between the previous hunk+nhunk
and the new block base returned by mysbrk() is usable.
the real problem is that gethunk() does not grow the allocation
but just allocates a new hunk, so repeated calls to gethunk()
wont make nhunk grow to satisfy the allocation.
The swcursor used a 32x32 image for saving/restoring
screen contents for no reason.
Add a doflush argument to swcursorhide(), so that
disabling software cursor with a double buffered
softscreen is properly hidden. The doflush parameter
should be set to 0 in all other cases as swcursordraw()
will flushes both (current and previours) locations.
Make sure swcursorinit() and swcursorhide() clear the
visibility flag, even when gscreen is nil.
Remove the cursor locking and just do everything within
the drawlock. All cursor functions such as curson(),
cursoff() and setcursor() will be called drawlock
locked. This also means &cursor can be read.
Fix devmouse cursor reads and writes. We now have the
global cursor variable that is only modified under
the drawlock. So copy under drawlock.
Move the pc software cursor implementation into vgasoft
driver, so screen.c does not need to handle it as
a special case.
most pc's are multiprocessors these days, that use apic or
msi interrupts, then the irq does not matter anymore. and
uefi does not even assign irq to pci devices anymore. if
we have a problem enabling an interrupt, we will print.
9bootpxe: continues the war against random DHCPv6 DUIDs
Some UEFI implementations use random UUID based DUID instead of
ethernet address, but use ethernet derived link-local addresses.
So extract the MAC from our IPv6 address.
memory returned by rampage() is not zeroed, so we have to
zero it ourselfs. apparently, this bug didnt show up as we
where zeroing conventional low memory before the new memory
map code. also rampage() never returns nil.
This replaces the memory map code for both pc and pc64
kernels with a unified implementation using the new
portable memory map code.
The main motivation is to be robust against broken
e820 memory maps by the bios and delay the Conf.mem[]
allocation after archinit(), so mp and acpi tables
can be reserved and excluded from user memory.
There are a few changes:
new memreserve() function has been added for archinit()
to reserve bios and acpi tables.
upareserve() has been replaced by upaalloc(), which now
has an address argument.
umbrwmalloc() and umbmalloc() have been replaced by
umballoc().
both upaalloc() and umballoc() return physical addresses
or -1 on error. the physical address -1 is now used as
a sentinel value instead of 0 when dealing with physical
addresses.
archmp and archacpi now always use vmap() to access
the bios tables and reserve the ranges. more overflow
checks have been added.
ramscan() has been rewritten using vmap().
to handle the population of kernel memory, pc and pc64
now have pmap() and punmap() functions to do permanent
mappings.
This is a generic memory map for physical addresses. Entries
can be added with memmapadd() giving a range and a type.
Ranges can be allocated and freed from the map. The code
automatically resolves overlapping ranges by type priority.
ether8003: use physical addresses for ISAConfig ether->mem
Fix the inconsistent use of ether->mem. Always use physical
addresses. Let ether8390 convert to virtual addresses using
KADDR() when we have to copy data in/out.
cinap_lenrek [Tue, 31 Mar 2020 19:57:53 +0000 (21:57 +0200)]
ip/dhcp6d: work arround non-ethernet based client duid
in ndb, we use the ethernet mac to identify the client.
in dhcpv6, there is just a uniqueue device id that
might even be generated randomly. to find the ethernet
address of a client, check the duid type and only use
it when the dudid is of type 1 (link layer) or 3 (link
layer address + time) and the link layer address type
is 1 (ethernet). otherwise, assume the source ip is
a link local address and extract it from that.
this hack works for thinkpad t495, which uses random
uuid based client duid.
cinap_lenrek [Sat, 28 Mar 2020 23:44:09 +0000 (00:44 +0100)]
pc: fix mp boot regression due mistake in mkfile
the previous mkfile had a sneaky hack that would use
sed to delete the first 2 lines of hex output to strip
the 32 byte long a.out header for apbootstrap and rebootcode.
use 8l -H3 flag to strip the header from the output file.
Alex Musolino [Thu, 26 Mar 2020 07:54:39 +0000 (18:24 +1030)]
grep: fix handling of -b flag
Output buffering is automatically disabled when reading from stdin.
In this case, supplying the -b flag ought to be redundant. However,
since Bflag was being XORed into the flag set - rather than simply
ORed - supplying -b would actually enable output buffering.
Ori Bernstein [Tue, 24 Mar 2020 21:45:15 +0000 (14:45 -0700)]
turn ptrdiff_t into a 64 bit type
while technically a 32 bit ptrdiff_t is in spec on
systems with 64 bit ponters as long as we guarantee
that individual objects are small enough, this can
confuse legitimate code, so lets fix this.
Ori Bernstein [Sun, 15 Mar 2020 22:08:04 +0000 (15:08 -0700)]
fix ccom idempotency
ccom may be called multiple times on the same
node, via 'goto loop' calls from the commute
label, OADD, and a few other places.
Casts to void could null out the LHS of the
node, which would cause the compiler to crash
if the cast was revisited due to one of these
cases, because we tried frobbing n->left.
Ori Bernstein [Tue, 10 Mar 2020 17:09:34 +0000 (10:09 -0700)]
improve usage messages (thanks henesy)
Fix inconsistencies between programs and their usage
messages, correct instances where information seems
to be missing or lost. This includes missing arguments,
making usage consistent with manuals, and so on.
Ori Bernstein [Mon, 9 Mar 2020 15:02:22 +0000 (08:02 -0700)]
fix alignment in ape malloc
We used to have a padding int in the structure
after the next pointer, to align it to 16 bytes.
On 64 bit architectures, the pointer was already
8 bits, so the padding misaligned things to 20
bytes.
This fixes it so that we're explcit about the
data alignment we want, instead of hoping that
the various sizes line up.
cinap_lenrek [Sun, 8 Mar 2020 21:00:23 +0000 (22:00 +0100)]
lib9p: implement automatic remove-on-close cleanup in postsharesrv(), remove postfd() and sharefd() functions
with the latest changes to shr(3), we can use ORCLOSE on
the control file to get the mount in the share automatically
removed when the server exits or something goes wrong during
postsharesrv().
do not expose postfd() and sharefd() functions. they where
undocumented and leak the control file descriptors.
cinap_lenrek [Sun, 8 Mar 2020 17:23:07 +0000 (18:23 +0100)]
devshr: unmount service on #σc/share/mount removal
when the control mountpoint side gets removed, close
mount channel immediately. this is usefull for implementing
automatic cleanup with ORCLOSE create mode.
cinap_lenrek [Sun, 8 Mar 2020 15:06:37 +0000 (16:06 +0100)]
nusb/usbd: cleanup processes on unmount
this makes sure that when postsharesrv() fails (for
example because the shr file already exists), the
worker process gets killed and all file descriptors
to devusb get closed.
cinap_lenrek [Sun, 8 Mar 2020 13:11:23 +0000 (14:11 +0100)]
devproc: return process id when reading /proc/n/ctl file
allow reading the control file of a process and return
its pid number. if the process has exited, return an error.
this can be usefull as a way to test if a process is
still alive. and also makes it behave similar to
network protocol directories.
another side effect is that processes who erroneously
open the ctl file ORDWR would be allowed todo so as
along as they have write permission and the process is
not a kernel process.
cinap_lenrek [Sun, 8 Mar 2020 04:52:23 +0000 (05:52 +0100)]
lib9p: restore previous behaviour of fd group sharing
it appears that too many fileservers rely on the fileserver
process sharing the filedescriptors with children of the
caller to postmntsrv() or threadpostmntsrv().