summaryrefslogtreecommitdiff
path: root/sys/arch
AgeCommit message (Collapse)Author
2002-11-10Use appropriate VM_MAP constants rather than hardcoded values inMiod Vallat
uvm_map_create() and uvm_km_suballoc().
2002-11-10some creamy fillingMichael Shalayeff
2002-11-09Instead of relying on uvm_useracc(), get a false sense of security, andMiod Vallat
do not check copyin() result, take care and properly handle copyin() failure. This was not harmful, but a bit more correctness never harms.
2002-11-09- make IFF_SIMPLEX only set if the media is IFM_FDX (prevents ip6 fromFederico G. Schwindt
seeing it's packets and thinking another machine with the same mac is on the wire when running half-duplex); from NetBSD. tested by matthieu, jason and him ok.
2002-11-08absolutely no need to save the fpu state before rewriting the fpu regs, just ↵Michael Shalayeff
reset the curpcb and force a reload
2002-11-08hmm, todos seem to be unapplicable no moreMichael Shalayeff
2002-11-08only reset the pending fpu exceptions on fork, not the rest of the fpu ↵Michael Shalayeff
status reg
2002-11-08prepare for bktr on macppc (commented out in generic for now); drahn@ okMichael Shalayeff
2002-11-08Don't uvm_useracc the user sigcontext in sys_sigreturn and then accessMichael Shalayeff
the user addresses directly from the kernel. copyin is faster and can correctly deal properly with mappings that uvm_useracc thinks are correct but will fault anyway (to figure out how to generate such mappings is left as en excercise for the reader).
2002-11-08Don't uvm_useracc the user sigcontext in sys_sigreturn and then accessMiod Vallat
the user addresses directly from the kernel. copyin is faster and can correctly deal properly with mappings that uvm_useracc thinks are correct but will fault anyway (to figure out how to generate such mappings is left as en excercise for the reader). Blatantly stolen from art@'s similar fix to sparc.
2002-11-07dev/isa/i82365_isasubr.c is needed for pcic at pci; from Bernard Gardner via ↵Michael Shalayeff
pr2962
2002-11-07pmap_[de]activate() are not nops; w/ art@'s helpMichael Shalayeff
2002-11-07make pte flushing into a separate function,Michael Shalayeff
add a missing pte flush in page_remove(), removed a few dead debugs and ifdefs. use local vars inestead of long -> chains./ miod@ ok
2002-11-07Don't uvm_useracc the user sigcontext in sys_sigreturn and then accessArtur Grabowski
the user addresses directly from the kernel. copyin is faster and can correctly deal properly with mappings that uvm_useracc thinks are correct but will fault anyway (to figure out how to generate such mappings is left as en excercise for the reader).
2002-11-07Small simplification in sendsig.Artur Grabowski
In sys_sigreturn - don't uvm_useracc the context address and then copyin it afterwards. The copyin deals better with finding if the user address is correct and it's faster.
2002-11-06- always initialize colormaps, even if the frame buffer is non console; thisMiod Vallat
helps if the ramdac does not get initialized (idea from jason@) - only register a shutdown hook for the frame buffers which need it, if this is the console frame buffer. Otherwise this is just a waste of time.
2002-11-06Eliminate the use of KERN_SUCCESS outside of uvm/Artur Grabowski
Also uvm_map returns KERN_* codes that are directly mapped to errnos, so we can return them instead of doing some attempt to translation. drahn@ "I see no problem" pval@ "makes sense"
2002-11-05Add a WEAK_ALIAS macro, to be used by libc/libc_r shortly.Miod Vallat
2002-11-05Allocate by default 6 virtual consoles on alpha, as on i386, rather thanMiod Vallat
only one. X11 is configured here to start on ttyC5, like on i386 again. This requires alpha-specific changes to the X servers, that went in the tree before 3.2 was released. On tga-compatible boards, there are no virtual consoles available, and these changes will have no effect.
2002-11-03Missing word in buffers size line.Miod Vallat
2002-11-01Working atomic locks on vax for libc_r.Miod Vallat
ok marc@
2002-11-01add a missing delay after printf on halt[-p], proper arg for a cold hookMichael Shalayeff
2002-10-31lasi on j210 is in a strange place, add an exact entry for now, to be maybe ↵Michael Shalayeff
rethought later once more such cases pile up; from millert@
2002-10-30as found on a 256m machine of millert@'s allocate pdes for the whole kernel ↵Michael Shalayeff
virtual and do not bother (just four anyway); as a side work do not adjust addr,size since there is no need
2002-10-29Since memory deallocation can't fail, remove the error return fromArtur Grabowski
uvm_unmap, uvm_deallocate and a few other functions. Simplifies some code and reduces diff to the UBC branch.
2002-10-29no need to restore shadowed reg, shaves 10% from runtime for tlb handler on ↵Michael Shalayeff
pcxl, more on pcxs/pcxt due to conditional pte ref update; tested on pcxt,pcxl by me on pcxs by miod
2002-10-29Remove some unused globlsArtur Grabowski
2002-10-28do not use asm for accessing the page tables since they are mapped now.Michael Shalayeff
fix ptp accounting and move diagnostic check in pmap_destroy() into a DIAGNOSTIC and it has not caught a one problem so far. when random-allocating the space ids use linear rehashing instead of a full new random which produces a better cache locality. miod@ ok
2002-10-28Convert sparc pmap from physseg to VM_PAGE_MD.Artur Grabowski
This allows us to remove some ambiguities on how some functions are called, remove some diagnostic checks for conditions that can never happen and remove the ugly hack with "pmap_initialized". It also removes some unnecessary overhead in time-critical functions like pmap_{zero,copy}_page and pmap_{is,clear}_{mod*,ref*}. miod@ ok
2002-10-27Turn into a really useful configuration file, i.e. GENERIC tuned for sun4cMiod Vallat
hardware, like SUN4M is for sun4m hardware.
2002-10-27Typo in commentMiod Vallat
2002-10-25fake intrcnt (for now) such that vmstat works (;Michael Shalayeff
2002-10-25allow reading the direct mapped segment through the kmemMichael Shalayeff
2002-10-24Add missing pmap_update(9) after pmap_kenter_pa(9); art@ ok.Federico G. Schwindt
2002-10-24nlist does no aout no more, fuget about som tooMichael Shalayeff
2002-10-24use loadfileMichael Shalayeff
2002-10-24Add va_copy() as necessary for powerpc, same protections as was addedDale Rahn
recently on other architectures.
2002-10-23Add a trivial va_copy() macro to all architectures but powerpcTodd C. Miller
(which I will leave for Dale since it needs special handling). From NetBSD (and same as sparc64). espie@ OK
2002-10-22do not force single-user; pt out by fries@Michael Shalayeff
2002-10-21make an ieeefp regress pass; miod@ okMichael Shalayeff
2002-10-21try harder to sync in dmamap_syncMichael Shalayeff
2002-10-20fxp now works on bigendian, add it to macppc.Dale Rahn
2002-10-19Kill register.Hugh Graham
Add intvec counting. Adapt to openbsd WOPEN handling. All untested but does compile...
2002-10-18this is a precision architecture -- be more precise about fault types and in ↵Michael Shalayeff
sigsegv deliveries; vm_ssize is in pages, apparently
2002-10-18Remove some very pointless 'if 0' code, some of which prevented compilationDale Rahn
with gcc3.2, and add a ';' for a case statement as it requests.
2002-10-18Do not define astpending twice, once non-volatile and one volatile.Dale Rahn
2002-10-17use shadows on fpu cxsw (and fix arg0 trashing), do not restore shadowed ↵Michael Shalayeff
regs before rfir
2002-10-17sync with gcc version, fixes bug in sftp where it had a 64 bit parameterDale Rahn
which straddled the last register first stack parameter.
2002-10-17Correct io space config after last patch. allows fxp to workDale Rahn
(with pending fxp BE diffs)
2002-10-17convert to use vm_page_md instead of pmap_physseg, make code smaller and ↵Michael Shalayeff
simpler, indeed; after art's suggestion and by looking into his diffs oneyed