| Age | Commit message (Collapse) | Author |
|---|
|
1. Add bounds checking to vga_pci_mmap, fixing a potential security issue.
Limit the amount of memory to what the vbe info structure says the card
card has. This should be nicely refined to limit to the amount of memory
needed by the active resolution and depth but this will do for the time
being.
2. Fix several places in vesafb.c where the return of kvm86_bios_addpage()
isnt being checked. Also ensure that if vesafb_get_mode_info() fails that it
cleans up after itself by releasing the page it added.
3. Correct the range checks in vesafb_putcmap and vesafb_getcmap, harmonize
code with similar code found else where.
ok miod@
|
|
From Brad, tested by Brad, biorn@ and Johan M:son Lindman.
|
|
|
|
the interface will fail to initialize with an EEPROM error if the interface
does not have a link upon boot.
Tested by mk@ and janek@
From Jack Vogel@Intel via brad
|
|
sizing its BARs.
tested niallo@, tsi@, ian@, mk@, sturm@, kettenis@, krw@, beck@, drahn@
ok beck@, kettenis@, drahn@
|
|
|
|
Identical fixes from myself and brad@, also reported by chefren@pi.net.
|
|
set up and allocated (which happens in a mountroothook). this prevents an
early call to the interrupt handler from causing a null deref when trying
to look into the unallocated regions.
found by mcbride when ciss and bnx were sharing an interrupt. mounting
root caused interrupts before the bnx was properly set up.
"commit your fix" mcbride@
|
|
|
|
changes but with the more conservative (i.e. specific) chip matching
logic of Linux's tg3. No change except to pci express chips.
From Brad.
Tested by brad & thib@ (BCM5750) and pedro la peu (BCM5752).
|
|
From Brad, tested by Brad, wilfried@ and Johan M:son Lindman.
ok wilfried@
|
|
eliminating a one line called once function by inserting its code
inline.
No functional change.
From Brad.
|
|
From brad@
|
|
From brad@
|
|
resolutions to attach without disturbing each other.
|
|
device is the system console; the interface currently passes the pci
attachment args and the sti region information, and will hopefully turn to
be flexible enough.
This allows sti@pci to be used as a console device on hppa, finally.
ok mickey@
|
|
when we need to access it; and change the rom parser to walk all the rom
images and pick the first one sporting code for the processor we run on,
instead of picking the first.
This allows dual-sti@pci setups to work.
ok mickey@, dual-head test martin@
|
|
sti_end_attach() if an error has occured.
ok mickey@
|
|
NetBSD.
|
|
THIS MEANS YOU NEED TO UPDATE YOUR FIRMWARE FILE BEFORE BOOTING WITH
A NEW KERNEL
tested by marco, biorn
|
|
|
|
|
|
|
|
Tested by many (IP/UDP/TCP):
Jason McIntyre <jmc@>
Chris Kuethe <chris.kuethe AT gmail.com>
Alf Schlichting <a.schlichting AT lemarit.com>
Rodolfo Gouveia <rgouveia AT cosmico.net>
Peter Stromberg <wilfried@>
Has been in snaps for weeks too with noone complaining so far.
|
|
Okay robert@
|
|
|
|
|
|
on all architectures, and have the driver cope with this as well
tested by ian
YOU MUST UPDATE YOUR MICROCODE FILE OR THE DRIVER WILL NOT BE HAPPY
|
|
From brad@
|
|
while im here.
|
|
contains random garbage on the stack.
|
|
the dva of the rfis and command list.
|
|
spec.
|
|
structs at all the right bits of it. the dma addresses arent taken care of
yet, just the kva ones.
|
|
|
|
continuing when busy flag set.
|
|
See similar commit to dev/usb/if_rum.c for more info. With this commit
all drivers have been switched away from the incorrect M_DUP_PKTHDR() usage.
OK mglocker@
|
|
|
|
first command byte is written to the card over the pci bus. the driver has
to capture this moment to get a precise timedelta. so make sure the code
fragment that takes the internal timestamp and sends the command byte can not
be interrupted.
this function is called with a low frequency (currently 0.1 Hz).
ok claudio, mglocker
|
|
change.
from brad@ ok mglocker@
|
|
|
|
|
|
instead of using static strings.
ok deraadt
|
|
Use these two functions in pci_devinfo() to remove some duplicated code and
make it easier to read. Remove two now unused variables. This shrinks the
size of the resulting object code.
Drivers (or other kernel code) can now get at the PCI vendor and/or product
strings using pci_findvendor() and pci_findproduct().
ok claudio, deraadt, mglocker
|
|
|
|
from brad@
|
|
timedelta sensor desc field and not the full description produced by
pci_devinfo() (which gets truncated in this case). There is no need to
have the PCi vendor string in the description as this is apparent from the
driver name already.
problem noticed by deraadt.
|
|
From brad@
|
|
hardware resets.
|
|
malloc'ed is completely bzeroed before use.
ok deraadt@
|
