Aren/NetBSD
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
110,293 Commits 606 Branches 0 Tags 3.1 GiB
e6b13790fa
Commit Graph

284 Commits

Author SHA1 Message Date
scw
e6b13790fa SH5 needs __HAVE_FUNCTION_DESCRIPTORS defined in order for ld.elf_so 
to resolve SHmedia function pointers correctly.
 2003-03-26 14:46:32 +00:00
scw
eefd7219fd Sign-extend the return value, as required by the ABI. 
While I'm here, simplify the asm operands/constraints.
 2003-03-26 12:00:41 +00:00
scw
955a468334 Add PIC support. 2003-03-24 14:26:16 +00:00
scw
2e5c2cd2ae Define just one mcount for both 32 and 64-bit ABIs; there's really no 
advantage in having two slightly different versions.
Also save/restore FP register parameters.
 2003-03-24 14:24:38 +00:00
scw
28db85d8f6 Relax the initial user stack alignment from 16 to 8 bytes. 
Add a workaround for what appears to be a bug in binutils whereby
the entry point of an SHmedia binary can sometimes specify
SHcompact code (LSB is clear).
 2003-03-24 14:21:27 +00:00
scw
5d74e4ddb2 Now that we can deal with debug exceptions, use a BRK instruction for 
ddb breakpoints instead of TRAPA in order to simplify the trapa (syscall)
handler.
 2003-03-19 11:37:57 +00:00
scw
2bce28091c Fix some comments. 
Set up stack frames in copyin/copyout and friends (if DDB is defined) so
that ddb stack traces can trace through them.
 2003-03-19 11:34:28 +00:00
scw
7c18481cf2 G/C an obsolete comment. 2003-03-19 11:31:55 +00:00
scw
bfca0494bd Try to configure all children of cprc0, not just clock. 2003-03-13 14:37:36 +00:00
scw
d4f824014d A bunch of things: 
- Clean up the way cpu-specific tlb/cache functions are configured
    and used.
  - Add a workaround for a problem whereby cpu* at superhyway? fails
    to probe.
  - Print more info about the cpu/cache.
  - Move the RESVEC handlers back into generic sh5 code and ditch
    the panic stack hack.
  - Make the on-chip SCIF device the default console on Cayman.
  - Add experimental support for booting via a standalone bootstrap
    program (not yet committed) and using the boot parameters passed
    in by it.
  - Add a few more SH elf constants.
  - Tick a couple of items off the TODO list.
 2003-03-13 13:44:16 +00:00
scw
31b3a7c37f In scif_getc(), check SCIF_SCFSR2_RDF in addition to SCIF_SCFSR2_DR to 
detect waiting Rx data. This fixes an occasional console lockup when
when pasting text at the ddb prompt.
 2003-03-13 13:19:01 +00:00
tshiozak
31e2cbf0b5 add some ISO C 1995 I18N functions and types: 
btowc, wctrans, towctrans, wcscoll, wcsxfrm, wctype_t and wctrans_t.
 2003-03-02 22:18:11 +00:00
kent
cd7d9faeaf Introduce BUS_DMA_NOCACHE, and bus_dmamem_map() of i386 supports it. 2003-01-28 01:07:51 +00:00
scw
e212eb9e09 Fix pasto in last commit. 2003-01-22 13:46:44 +00:00
scw
a4b020d99e In cpu_upcall(), validate the address of the upcall handler so as to 
avoid a kernel-mode exception if restoring an invalid SPC.
 2003-01-22 13:40:57 +00:00
scw
3b0144945a Fix a cpu_switchto() botch. 2003-01-22 13:38:51 +00:00
scw
fe8c97c2e1 In _MCONTEXT_TO_SIGCONTEXT(), set r24 to the magic number expected 
by sh5's sys___sigreturn14().
 2003-01-22 13:38:11 +00:00
scw
1ec8a7e381 Add _UC_UCONTEXT_ALIGN. 2003-01-22 13:35:55 +00:00
scw
2dd1c1657f Add sigcontext/mcontext conversion macros. 2003-01-21 11:28:01 +00:00
scw
a57f525335 Define {,UN}LINK_BIG_FRAME() macros for dealing with stackframes 
greater than 511 bytes in size.
 2003-01-21 11:27:18 +00:00
scw
c3be47693c Use 'int' instead of 'u_int32_t' in a couple of places to avoid 
depending on sys/types.h.
 2003-01-21 11:26:04 +00:00
scw
0767992461 If we don't explicitly recognise the trap type, only panic if it 
came from kernel mode. Otherwise, print details of the exception
and send a SIGILL to the process.

The is necessary now that debug exceptions are handled here.
 2003-01-20 22:32:26 +00:00
scw
def0257e54 Split struct reg into integer and FP parts, to be more like every other port. 
As a side benefit, advertise PT_[GS]ETFPREGS.
 2003-01-20 20:07:52 +00:00
scw
c64b785002 Add sh5 kernel support for SAs. 
This gets the Cayman to multi-user with a non-SA aware userland.
 2003-01-19 19:49:47 +00:00
scw
8b92aecb3f Use the full 'sh5le-unknown-netbsd-' prefixed toolchain names, and 
set optimisation to -O2 now that gcc seems to generate decent sh5 code.
 2003-01-19 19:46:57 +00:00
wiz
1035faff1d writable, not writeable. 2003-01-06 20:30:28 +00:00
wiz
c6f1277bfa synchronous, not syncronous. 2003-01-06 13:10:25 +00:00
thorpej
dca15fc8c2 Use aprint_normal() for cfprint routines. 2003-01-01 02:10:08 +00:00
manu
4a06119a9d Pass the system call table to trace_enter() and ktrsys() so that it is 
possible to use alternate system call tables. This is usefull for
displaying correctly the arguments in Mach binaries traces.

If NULL is given, then the regular systam call table for the process is used.
 2002-12-21 16:23:56 +00:00
thorpej
e8cc3884de Rename __LDPGSZ to AOUT_LDPGSZ, to accurately reflect what it is. 2002-12-10 17:14:02 +00:00
thorpej
78ea2dd367 Use __LDPGSZ (which must be == USRTEXT) as the text address for a.out 
executables, and eliminate the USRTEXT constant, which was only used
by the a.out exec code.
 2002-12-10 05:14:24 +00:00
scw
d4cfbd3754 Implement __cpu_simple_lock{,init,try}() for sh5. 2002-12-07 22:31:33 +00:00
scw
50247a2a3c Fix PRIiPTR. 2002-12-07 21:37:35 +00:00
scw
57c2c2b87f Add a hack at the end of setregs() which works around a bug which can, 
under some circumstances, leave turds in the icache following vmspace
teardown.

It's not yet clear if this is a pmap bug or a toolchain problem since
the hack is unecessary when the kernel is compiled with -O0. Of course
that could just be masking the problem due to increased icache pressure...
 2002-12-06 10:22:15 +00:00
scw
f051850107 Avoid strict-alias warnings. 2002-12-06 10:14:10 +00:00
scw
41085b101b fsina.s, fsrra.s and fcosa.s take single-precision operands, not double. 2002-12-06 10:12:13 +00:00
scw
87897c05fe Qualify ptel and tlbcookie fields with volatile since they are 
modified by the TLB miss handler.
 2002-12-06 10:10:48 +00:00
lukem
0635de35a3 Remove KDIR=, since SYS_INCLUDE=symlinks and KDIR are not supported any more. 2002-11-26 23:30:07 +00:00
scw
918d34d917 Quell an uninitialised variable warning. 2002-11-24 12:07:52 +00:00
scw
2fcea657de Implemented #2: Improved TLB management 2002-11-23 09:27:03 +00:00
scw
a3885e9bdd - Change VM_MIN_KERNEL_ADDRESS to start at KSEG1 instead of KSEG0. 
- Overhaul the TLB management code such that we now keep track of
   the exact TLB slot at which a mapping was inserted, both for user-
   space and kernel mappings. This addresses #2 on the TODO list.
 2002-11-23 09:25:54 +00:00
scw
dbaff62d21 Make this compile again. 2002-11-23 09:14:14 +00:00
manu
d584ed9598 Add a realcode argument to trace_enter and ktrsyscall. realcode is the 
original system call number, which can be negative for a Mach trap.
We cannot just replace code by realcode, because ktrsyscall uses it as
an index in the system call table, thus crashing the kernel when the
value is negative.
 2002-11-15 20:06:00 +00:00
scw
bc159224ed Missing newline. 2002-10-31 14:52:54 +00:00
scw
78aeffb90b Add an explicit ULL suffix to a 64-bit constant to appease gcc-current. 2002-10-31 14:52:30 +00:00
scw
3795c4e6f4 Since we keep a list of all mappings for a given physical page (even 
unmanaged mappings) so we can deal with cache aliases, make sure to
skip unmanaged/wired mappings (added via pmap_kenter_pa()) when doing
things like pmap_page_protect().
 2002-10-31 14:34:17 +00:00
scw
6a3f09110b Catch NMIs when DDB is defined, instead of DEBUG. 
Drop the interrupt level to zero when before handling ASTs in the
exit path of h/w interrupts.
 2002-10-31 14:24:16 +00:00
scw
8bf915f645 When setting the SP for a process, copy the same value to the 
process' FP for the sake of consistency.
 2002-10-31 14:20:39 +00:00
jdolecek
c82ab2eb79 now that mem_no is emitted by config(8), there is no reason to keep 
copy of more or less identical iskmemdev() for every arch; move the function
to spec_vnop.c, and g/c machine-dependant copies
 2002-10-26 13:50:17 +00:00
scw
b67ad48827 Put bit definitions for the FP control/status register in <sh5/fpu.h> and 
use them instead of "magic" numbers in the source.
 2002-10-24 13:58:48 +00:00