The boot blocks do a primitive form of autoconfiguration. Duplicate some
of the old config strcutures for the benefit of the boot blocks; they
don't need something as flexible as the kernel does.
- Always ensure that RAW_PART ("c") can be opened.
- Disallow unused or non-existent partitions from being opened.
- Don't do bounds checking or partition translation on RAW_PART.
This is consistent with other disk drivers in the NetBSD source tree,
and fixes a condition where the user could not fix a corrupted disklabel
due to a bogus offset for partition "c".
- Always allow RAW_PART to be opened, regardless of the partition table.
- Never do partition translation on RAW_PART; it's always offset 0.
- Always make sure the disklabel is read from RAW_PART.
- Make *strategy() return values consistent.
This fixes a condition where a bogus partition table could be written
to the disk from SYS_INST, and the user had no way to rewrite with
a correct partition table, since RAW_PART would be invalid.
We're about 75% there. SCSI and HP-IB are not yet supported in a new
config kernel; some autoconfiguration hackery has to be done there, yet.
These changes are enough to network boot a diskless kernel.
New config glue is enabled with the "NEWCONFIG" kernel option. If that
option is not present, an old config kernel will be built. Any kernel
configured with config(8) will automatically pick up the NEWCONFIG
option from std.hp300.
dmacomputeipl(), which is called by drivers which use DMA once they've
hooked up their interrupts. This new function computes the appropriate
ipl to use for the DMA controller and (re-)establishes it's interrupt.
on indirect-config busses a (permanent) softc that they could share
between 'match' and 'attach' routines:
Define __BROKEN_INDIRECT_CONFIG so that old autoconfiguration
interfaces are used, until drivers are converted to use the new
interfaces (actually, converted back to use the _older_ interfaces)
which prohibit indirect configuration devices from receiving a softc
in their match routine that they can share with their attach routine.
Lets users over-ride with makeoptions COPTS="..." in kernel config files.
Leave `mandatory' flags (like -msoft-float which on m68k enforces no
FP in kernel) in CFLAGS.
VM_PROT_READ|VM_PROT_EXECUTE. The previous default (VM_PROT_ALL)
would cause the following scenario:
- someone attempts to write kernel text (my test was writing
to an offset of /dev/kmem which was known to be in the text
segment, while in single-user mode).
- enter trap() with MMU fault (because of RO pte).
- trap() calls vm_fault(), which looks up vm_map_entry for
fauling address.
- vm_fault interprets write fault and VM_PROT_WRITE (in VM_PROT_ALL)
as COW; new page allocated, data copied to new page, new page
mapped in at trunc_page(<faulting va>).
- wow, look at the fireworks!
Fixes two potential symptoms:
- kernacc() returns TRUE when checking for permission to write
an offset in kernel text, which is bogus, since the text has
been mapped RO by pmap_bootstrap().
- Handling of a stray pointer that attempted to scribble into
kernel text would not be executed properly.
don't do this, kernacc() will bogusly return TRUE for page 0, causing a
NULL pointer dereference in uiomove() when reading /dev/kmem.
Thanks to Scott Reynolds for noticing the problem.
- Fix up usage of MBD_ISPID().
- Ensure that we never deal with a NULL proc, and if our proc has
no pcb, punt. Suggested by Gordon Ross.
- Eliminate some redundant NULL pointer checks in the T_MMUFLT cases;
proc0.p_addr is now initialized early, and we make a single test
for sanity at the top of trap() now.
- Initialize proc0.p_addr just after setting up the kernel stack, to avoid
getting NULL pointers in trap(). Change suggested by Gordon Ross.
- Panic if main() returns.
- Tidy up a couple of comments.
(remaps page read/write/cache-inhibit, does write, restores previous
mapping). Kernel text no longer needs to be read/write with DDB/KGDB
is in the kernel.
Based on a similar module written for the Sun3 port by Gordon Ross,
and modified somewhat by me.
- Offset kernel text one page. Stash the PA of this offset page for
use later.
- Add a few comments.
- Free up some registers earlier in the initialization process.
- Use a `prototc' to set the Translation Control register, rather
than relying on a2 pointing just past the MMU trampoline in the
high page. (Suggested by Charles Hannum.)
- Set VBR to the kernel vector table just before turning on the MMU.
- Just before rebooting, set VBR to 0, which is what the BOOTROM expects
it to be.
kernel setroot(). The device type for network boot is set like any
other device now. Also, call the punit entry point for device drivers
to properly set `bootdev'.
The "options GENERIC" entry in kernel config files is not longer necessary
for "swap generic" kernels. Uses new config constructs which work with
some glue in an old config environment. This code will support new config
with minimal changes.
an unrecognized keyboard produced garbage on keypresses in the kernel,
but worked in the boot program (which has this default). The bug is
that the keymap pointers are unitialized.
- Only pass user trace traps and breakpoints on to trap().
Gets rid of some hair in the trace/breakpoint trap cases.
- Before entering the debugger, switch to a temporary
stack so that the debugger can alter the stack pointer.
- Add glue for KGDB (still not complete).
Some other minor cleanup:
- Protect against some bad pointer derefs.
- Be more a little more verbose when a fatal trap
occurs to aid debugging.
- Only pass user trace traps and breakpoints on to trap().
- Before entering the debugger, switch to a temporary
stack so that the debugger can alter the stack pointer.
- Add glue for KGDB (still not complete).
Clearly mark the MMU enable trampoline code.
thorpej