a5fc9c0bbe
Use physical addresses at the interface level, letting drivers remap
...
them as appropriate.
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:35 +01:00
7d7ee22121
Prevent 64-bit constants from being cropped to 32 bits when used in C code.
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:34 +01:00
2c93e12cfe
Avoid tlbw* hazards for the R4600/R4700/R5000.
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:34 +01:00
c3455b0efc
Inline ioremap() calls for constant addresses that map to KSEG1.
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:34 +01:00
c134a5ecdb
Avoid defining variables in the middle of a block which breaks older
...
compilers.
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:33 +01:00
7222424e2e
More .set to keep 32-bit processors happy.
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:33 +01:00
4c0a2d4275
Fix the diagnostic dump for the XTLB refill handler.
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:32 +01:00
41986a6e7e
Fix a diagnostic message.
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:32 +01:00
362b1d546d
Conversion to plat_setup() for TX4927 also.
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:32 +01:00
01d42abdf3
Using get_nasid() to find the console node will blow up nicely if
...
called on a another node than the console node, so use the master_nasid
instead and in the unlikely case that one isn't initialized yet, fall
back to get_nasid().
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:31 +01:00
c4559f67b7
Always use ".set mips3" rather than select between "mips2" or "mips3"
...
for assembling ll/sc sequences to avoid problems with 64-bit
configurations.
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:31 +01:00
69c75fb458
Actual handlers for bus errors for Pmax and 3min.
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
diff --git a/arch/mips/dec/kn01-berr.c b/arch/mips/dec/kn01-berr.c
new file mode 100644
2005-10-29 19:31:30 +01:00
64dac503e8
System-specific handling of bus errors for DECstation variations
...
supporting parity errors only for memory (Pmax/3min/Maxine).
Fixes for resources decoded by the KN04/KN05 MB ASIC. Additional
clean-ups for the ECC handler.
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:30 +01:00
3b2396d972
Use correct names for bits in the R3k cp0.status register.
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:29 +01:00
e20368d5df
Get the thing to compile again ...
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:28 +01:00
c6ad7b7d3c
Use macros for the RM7k cp0.config bits instead of magic numbers.
...
Minor clean-ups.
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:28 +01:00
8a185d14b6
Fix types for firmware arguments. Don't define unneeded messages.
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:27 +01:00
2e2849670a
Remove left-over unused bits.
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:27 +01:00
260c96738c
Mark __die() "noreturn" for real.
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:26 +01:00
778220f7c4
Fix function types to ones appropriate for initcalls.
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:24 +01:00
a76f9fe122
GCC 4.0.0 broke `attribute(("alias"))' -- resort to an assembly variant.
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:24 +01:00
3bd4c902da
Deal with the bloody KSEG vs CKSEG horror...
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:23 +01:00
902d21d531
There is NO port I/O space on the DECstation. Minor clean-ups.
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:23 +01:00
02416dcf5a
Redo RM9000 workaround which along with other DSP ASE changes was
...
causing some headache for debuggers knowing about signal frames.
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:23 +01:00
aac8aa7717
Enable a suitable ISA for the assembler around ll/sc so that code
...
builds even for processors that don't support the instructions.
Plus minor formatting fixes.
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:22 +01:00
fded2e508a
Optimize R3k TLB Load/Store/Modified handlers, by scheduling
...
delay slots properly and avoiding an unnecessary jump to a jump.
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:22 +01:00
d925c262dd
Fill R3k load delay slots properly.
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:21 +01:00
9678e28b1a
Only dump instructions actually emitted.
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:21 +01:00
15b6e09b66
dec_esp: Use physical addresses
...
These should really be addresses obtained with ioremap() or some
bus-specific backend, but for now...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:20 +01:00
68e4a86c80
This interrupt is *always* handled -- MIPS_BE_DISCARD just means
...
no further action wanted.
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:20 +01:00
a93500037e
Fix dependencies for DECstation framebuffers.
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:19 +01:00
478489dd2c
Remove dead code which was causing warnings.
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:19 +01:00
d5b6f1db5d
For MIPS32/MIPS64 cp0.config.mt == 1 implies a standard (R4k-style)
...
TLB, so no need to set it separately for each implementation.
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:19 +01:00
81731f7997
The DbAu1500 board also support big endian. Gee, imagine that.
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:18 +01:00
ac130ac494
Fix build with CONFIG_PRINTK disabled.
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:18 +01:00
e50c0a8fa6
Support the MIPS32 / MIPS64 DSP ASE.
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:17 +01:00
10f650db1b
64-bit fixes for Alchemy code ;)
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:17 +01:00
ffd099bd33
Fix build for CONFIG_BUG=n. Yes, bugs are now a compile time option ;-)
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:17 +01:00
149f60b30c
When building for Atlas, Malta or SEAD convert the kernel to srecs by
...
default.
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:16 +01:00
f8280c8d3d
Fix tasteless #ifdef mess in audit_arch(), minor cleanups.
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:16 +01:00
fdb551a4c5
Bugs are now a configuration option.
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:15 +01:00
629c83f89b
On MIPS the struct sigev preamble is only 8 bytes.
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:15 +01:00
4a99d1e25b
Now that a struct is the only member left in struct
...
mips_fpu_emulator_stats cleanup that unnecessary nesting of structs.
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:14 +01:00
baee502ce2
Get rid of the eir struct mips_fpu_emulator_private member. It's
...
never initialized been initialized anywhere, just saved to and
restored from signal frames so nonsense anyway. As neat side effect
of being shared between all processors it was also abusable as a
nice covert channel between processes.
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:14 +01:00
1d74f6bc85
__compute_return_epc() uses CFC1 instruction which might result in a
...
coprocessor unusable exception since the process can lose its fpu
context by preemption.
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:13 +01:00
d547c5cc21
sys_nfsservctl() needs translation.
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:13 +01:00
b382fe8483
No point in checking cpu_has_tlb before we've computed the CPU options.
...
So for now we just unconditionally set the option - Linux wouldn't
work without a TLB anyway.
Setting MIPS_CPU_4KTLB was missing for Alchemy and Sandcraft, add that
back.
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:13 +01:00
4194318c39
Cleanup decoding of MIPSxx config registers.
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:12 +01:00
cd21dfcfbb
Fix preemption and SMP problems in the FP emulator code.
...
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:12 +01:00
63b2d2f4d2
Handle _PAGE_DIRTY correctly for CONFIG_64BIT_PHYS_ADDR on 32bit CPUs.
...
Signed-off-by: Thiemo Seufer <ths@networkno.de>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2005-10-29 19:31:11 +01:00
Maciej W. Rozycki