summaryrefslogtreecommitdiff
path: root/sys/arch/m88k/include
AgeCommit message (Collapse)Author
2024-10-23Rename `ci_intrdepth' to `ci_idepth'.Martin Pieuchot
ok jsg@, aoyama@
2024-10-22put opening { on same line as struct nameJonathan Gray
ok claudio@
2024-08-08Make user ci_flags is declared volatile.Miod Vallat
2024-07-14Add elf_aux_info(3)Jeremie Courreges-Anglas
Designed to let userland peek at AT_HWCAP and AT_HWCAP2 using an already existing interface coming from FreeBSD. Headers bits were snatched from there. Input & ok kettenis@ libc bump and sets sync will follow soon
2024-06-26return type on a dedicated line when declaring functionsJonathan Gray
ok mglocker@
2024-06-20Enable uvm percpu caches on luna88k.Kenji Aoyama
It survives 3.5 days "make build" and makes about 1.5% faster on 3 CPU machine:-) ok miod@ phessler@ dlg@
2024-06-09Add a compiler barrier where missing in CPU_BUSY_CYCLE() implemsJeremie Courreges-Anglas
Having differences between architectures is asking for problems. And adding a barrier here just makes sense in most cases. This is also what cpu_relax() provides in Linux land. ok kettenis@ claudio@
2024-05-28cpu_emergency_disable() was only used by mvme88k and can be removed now.Claudio Jeker
This removes one of the SCHED_LOCK usages in arch. OK miod@
2024-05-16Stop treating IPL_NONE mutexes special in MULTIPROCESSOR kernels. WhileMark Kettenis
in theory these are safe to use in code that runs under the kernel lock they are nasty trips when converting code to run without the kernel lock. ok mpi@, claudio@
2024-03-31Stop referring to dead-since-30-years gatherstats() in comments.Miod Vallat
2024-02-25clockintr: rename "struct clockintr_queue" to "struct clockqueue"Scott Soule Cheloha
The code has outgrown the original name for this struct. Both the external and internal APIs have used the "clockqueue" namespace for some time when operating on it, and that name is eyeball-consistent with "clockintr" and "clockrequest", so "clockqueue" it is.
2024-02-14Remove never ever used cpu_exec() macro, apparently already obsolete by theMiod Vallat
time this file was introduced close to 30 years ago.
2024-02-14Make sure pmap_unmap_direct flushes dirty cache lines.Miod Vallat
2024-01-24clockintr: switch from callee- to caller-allocated clockintr structsScott Soule Cheloha
Currently, clockintr_establish() calls malloc(9) to allocate a clockintr struct on behalf of the caller. mpi@ says this behavior is incompatible with dt(4). In particular, calling malloc(9) during the initialization of a PCB outside of dt_pcb_alloc() is (a) awkward and (b) may conflict with future changes/optimizations to PCB allocation. To side-step the problem, this patch changes the clockintr subsystem to use caller-allocated clockintr structs instead of callee-allocated structs. clockintr_establish() is named after softintr_establish(), which uses malloc(9) internally to create softintr objects. The clockintr subsystem is no longer using malloc(9), so the "establish" naming is no longer apt. To avoid confusion, this patch also renames "clockintr_establish" to "clockintr_bind". Requested by mpi@. Tweaked by mpi@. Thread: https://marc.info/?l=openbsd-tech&m=170597126103504&w=2 ok claudio@ mlarkin@ mpi@
2023-12-14NKMEMPAGES_MAX_DEFAULT is no longer used. Remove it from param.h.Claudio Jeker
OK miod@
2023-12-11Implement per-CPU caching for the page table page (vp) pool and the PTEMark Kettenis
descriptor (pted) pool in the arm64 pmap implementation. This significantly reduces the side-effects of lock contention on the kernel map lock that is (incorrectly) translated into excessive page daemon wakeups. This is not a perfect solution but it does lead to significant speedups on machines with many CPU cores. This requires adding a new pmap_init_percpu() function that gets called at the point where kernel is ready to set up the per-CPU pool caches. Dummy implementations of this function are added for all non-arm64 architectures. Some other architectures can probably benefit from providing an actual implementation that sets up per-CPU caches for pmap pools as well. ok phessler@, claudio@, miod@, patrick@
2023-07-25statclock: move profil(2), GPROF code to profclock(), gmonclock()Scott Soule Cheloha
This patch isolates profil(2) and GPROF from statclock(). Currently, statclock() implements both profil(2) and GPROF through a complex mechanism involving both platform code (setstatclockrate) and the scheduler (pscnt, psdiv, and psratio). We have a machine-independent interface to the clock interrupt hardware now, so we no longer need to do it this way. - Move profil(2)-specific code from statclock() to a new clock interrupt callback, profclock(), in subr_prof.c. Each schedstate_percpu has its own profclock handle. The profclock is enabled/disabled for a given CPU when it is needed by the running thread during mi_switch() and sched_exit(). - Move GPROF-specific code from statclock() to a new clock interrupt callback, gmonclock(), in subr_prof.c. Where available, each cpu_info has its own gmonclock handle . The gmonclock is enabled/disabled for a given CPU via sysctl(2) in prof_state_toggle(). - Both profclock() and gmonclock() have a fixed period, profclock_period, that is initialized during initclocks(). - Export clockintr_advance(), clockintr_cancel(), clockintr_establish(), and clockintr_stagger() via <sys/clockintr.h>. They have external callers now. - Delete pscnt, psdiv, psratio. From schedstate_percpu, also delete spc_pscnt and spc_psdiv. The statclock frequency is not dynamic anymore so these variables are now useless. - Delete code/state related to the dynamic statclock frequency from kern_clockintr.c. The statclock frequency can still be pseudo-random, so move the contents of clockintr_statvar_init() into clockintr_init(). With input from miod@, deraadt@, and claudio@. Early revisions cleaned up by claudio. Early revisions tested by claudio@. Tested by cheloha@ on amd64, arm64, macppc, octeon, and sparc64 (sun4v). Compile- and boot- tested on i386 by mlarkin@. riscv64 compilation bugs found by mlarkin@. Tested on riscv64 by jca@. Tested on powerpc64 by gkoehler@.
2023-07-02all platforms, kernel: remove __HAVE_CLOCKINTR symbolScott Soule Cheloha
Every platform made the clockintr switch at least six months ago. The __HAVE_CLOCKINTR symbol is now redundant. Remove it. Prompted by claudio@. Link: https://marc.info/?l=openbsd-tech&m=168826181015032&w=2 "makes sense" mlarkin@
2023-04-13pmap_copy() has never, ever, been implemented in any of the platforms OpenBSDMiod Vallat
ever ran on, and it's unlikely to ever be implemented, so remove it. ok jsg@
2023-02-12m88k also needs to define PMAP_CHECK_COPYIN as 1.Kenji Aoyama
"this makes sense" miod@
2023-01-31On systems without xonly mmu hardware-enforcement, we can still mitigateTheo de Raadt
against classic BROP with a range-checking wrapper in front of copyin() and copyinstr() which ensures the userland source doesn't overlap the main program text, ld.so text, signal tramp text (it's mapping is hard to distinguish so it comes along for the ride), or libc.so text. ld.so tells the kernel libc.so text range with msyscall(2). The range checking for 2-4 elements is done without locking (because all 4 ranges are immutable!) and is inexpensive. write(sock, &open, 400) now fails with EFAULT. No programs have been discovered which require reading their own text segments with a system call. On a machine without mmu enforcement, a test program reports the following: userland kernel ld.so readable unreadable mmap xz unreadable unreadable mmap x readable readable mmap nrx readable readable mmap nwx readable readable mmap xnwx readable readable main readable unreadable libc unmapped? readable unreadable libc mapped readable unreadable ok kettenis, additional help from miod
2022-12-06_C_LABEL() and _ASM_LABEL() are no longer useful in the "everythingPhilip Guenther
is ELF" world. Eliminate use of them in m88k code. ok aoyama@
2022-12-06m88k, luna88k: switch to clockintrScott Soule Cheloha
- Initialize tick_nsec during cpu_initclocks() We have no control over the interrupt clock on luna88k, so this switch is trivial. Bringup help and testing from aoyama@ and miod@. Link: https://marc.info/?l=openbsd-tech&m=166776371203450&w=2 ok aoyama@ mlarkin@
2022-09-10Remove pmap_collect() when a no-op, define __HAVE_PMAP_COLLECT otherwise.Miod Vallat
Use that define to shunt uvm_swapout_threads(), which is a noop when pmap_collect() does nothing. ok mpi@
2022-08-29use ansi volatile keyword, not __volatileJonathan Gray
ok miod@ guenther@
2021-08-30Remove typedef of db_addr_t; mpi converted the users of it to vaddr_t alreadyJasper Lievisse Adriaanse
back in 2019. ok mpi@
2021-07-06Introduce CPU_IS_RUNNING() and us it in scheduler-related code to preventMark Kettenis
waiting on CPUs that didn't spin up. This will allow us to spin down CPUs in the future to save power as well. ok mpi@
2021-06-02kernel: introduce per-CPU panic(9) message bufferscheloha
Add a 512-byte buffer (ci_panicbuf) to each cpu_info struct on each platform for use by panic(9). The first panic on a given CPU writes its message to this buffer. Subsequent panics on a given CPU print the panic message to the console but do not modify the buffer. This aids debugging in two cases: - If 2+ CPUs panic simultaneously there is no risk of garbled messages in the panic buffer. - If a CPU panics and then the operator causes a second panic while using ddb(4), the operator can still recall the first failure on a particular CPU. Misc. changes to support this bigger change: - Set panicstr atomically to identify the first CPU to reach panic(). - Tweak db_show_panic_cmd() to print all panic messages across all CPUs. Prefix the first panic with an asterisk ('*'). - Prefer db_printf() to printf() during a panic if we have it. Apparently it disturbs less global state. - On amd64, tweak fault() to write the local panic buffer. This needs more work. Prompted by bluhm@ and deraadt@. Mostly written by deraadt@. Discussed with bluhm@, deraadt@ and kettenis@. Borne from a discussion on tech@ about making panic(9) more MP-safe: https://marc.info/?l=openbsd-tech&m=162086462316143&w=2 ok kettenis@, visa@, bluhm@, deraadt@
2021-05-04Fix 2nd and 3rd arguments' cast in atomic_cas_ptr().Kenji Aoyama
This fixes compile errors (actually warnings) on m88k in sys/net/pf.c revision 1.1116 changes. Diff from Miod Vallat, tested on GENERIC and GENERIC.MP by me.
2020-09-25miod notes T_USER needs to be a bit, it was not.Theo de Raadt
The "snowflake" uniqueness of every MD trap impl often gets in the way of precisely & correctly interfacing to MI layers. The differences also complicates review, and causes new MI requirements to be incorrectly written. Thus an architecture will fall behind, not just because they are slow or rare, but because the code behaviour becomes increasingly incorrect. It is sad.
2020-07-08do not need these versions of timetc.hTheo de Raadt
2020-07-06Add support for timeconting in userland.Paul Irofti
This diff exposes parts of clock_gettime(2) and gettimeofday(2) to userland via libc eliberating processes from the need for a context switch everytime they want to count the passage of time. If a timecounter clock can be exposed to userland than it needs to set its tc_user member to a non-zero value. Tested with one or multiple counters per architecture. The timing data is shared through a pointer found in the new ELF auxiliary vector AUX_openbsd_timekeep containing timehands information that is frequently updated by the kernel. Timing differences between the last kernel update and the current time are adjusted in userland by the tc_get_timecount() function inside the MD usertc.c file. This permits a much more responsive environment, quite visible in browsers, office programs and gaming (apparently one is are able to fly in Minecraft now). Tested by robert@, sthen@, naddy@, kmos@, phessler@, and many others! OK from at least kettenis@, cheloha@, naddy@, sthen@
2020-06-30Remove obsolete <machine/stdarg.h> header. Nowadays the varargVisa Hankala
functionality is provided by <sys/stdarg.h> using compiler builtins. Tested in a ports bulk build on amd64 by naddy@ OK naddy@ mpi@
2020-05-31introduce "cpu_rnd_messybits" for use instead of nanotime in dev/rnd.c.David Gwynne
rnd.c uses nanotime to get access to some bits that change quickly between events that it can mix into the entropy pool. it doesn't use nanotime to get a monotonically increasing set or ordered and accurate timestamps, it just wants something with bits that change. there's been discussions for years about letting rnd use a clock that's super fast to read, but not necessarily accurate, but it wasn't until recently that i figured out it wasn't interested in time at all, so things like keeping a fast clock coherent between cpu cores or correct according to ntp is unecessary. this means we can just let rnd read the cycle counters on cpus and things will be fine. cpus with cycle counters that vary in their speed and arent kept consistent between cores may even be desirable in this context. so this is the first step in converting rnd.c to reading cycle counter. it copies the nanotime backend to each arch, and they can replace it with something MD as a second step later on. djm@ suggested rnd_messybytes, but we landed on cpu_rnd_messybits. thanks to visa for his eyes. ok deraadt@ visa@ deraadt@ says he will help handle any MD fallout that occurs.
2020-05-27Retire <machine/varargs.h>.Visa Hankala
Nothing uses the header anymore. OK deraadt@ mpi@
2020-05-26Rewrite m88k mutex code as a slight variation of the MI mutex code.Kenji Aoyama
This will make mutex spinning time visible in top(1), and also might improve stability. The major change in this is that the old assembly code acquires mutexes with an atomic exchange operation, but releases them with a regular store, but the new code always uses atomic exchange operations. The mutex.h changes to the macros conform to <sys/mutex.h> to be able to reset the system while in ddb. Suggested from Miod Vallat, tested by me. The stability in heavy load is greatly improved in my case.
2019-12-11%r26 hasn't been used by the threads implementation since 2016, so stopPhilip Guenther
resetting it in child_return() and update the comment in tcb.h to reflect reality ok miod@ aoyama@
2019-08-28sys/arch/luna88k/luna88k/locore0.S:dummy_cpu assumes thatKenji Aoyama
ci_mp_atomic_{begin,end} are 6th and 7th elements of cpu_info structure. Actually that is dummy structure used in early boot stage, but for consistency, move ci_srp_hazards position in cpu_info. ok mpi@
2018-12-05Include srp.h where struct cpu_info uses srp to avoid erroring out whenJonathan Gray
including cpu.h machine/intr.h etc without first including param.h when MULTIPROCESSOR is defined. ok visa@
2018-09-14Unify and bump some of the NMBCLUSTERS defines. Some archs had it set toClaudio Jeker
4MB which is far too low especially when the platform is able to run MP. New limits are, amd64 = 256M; arm64, mips64, sparc64 = 64M; alpha, arm, hppa, i386, powerpc = 32M; m88k, sh = 8M Still rather conservative numbers but much better than before. At least some hangs of arm64 build boxes was caused by this. OK kettenis@, visa@
2018-08-22Adding membar_xxx defines to userland.Kenji Aoyama
The src/lib/libc/thread/rthread.c 1.8 change adds #include <sys/atomic.h> in userland code. Current m88k atomic.h contents are inside of #if defined(_KERNEL) guard, then, nothing is defined for userland program. So we need adding some defines to compile it on m88k. The original diff is suggested from Miod Vallat, modified by the advice from mpi@ and kettenis@. ok kettenis@
2018-08-20Remove unused spllock().Visa Hankala
OK deraadt@ mpi@
2018-03-12Make luna88k use the MI mplock implementation.Kenji Aoyama
ok visa@
2018-03-05#define _MAX_PAGE_SHIFT in MD _types.h as the maximum pagesize an archTheo de Raadt
needs (looking at you sgi, but others required this before). This is for the circumstances we need pagesize known at compile time, not getpagesize() runtime. Use it for malloc storage sizes, for shm, and to set pthread stack default sizes. The stack sizes were a mess, and pushing them towards page-aligned is healthy move (which will also be needed by the coming stack register checker) ok guenther kettenis, discussion with stefan
2018-01-13Define and use IPL_MPFLOOR in our common mutex implementation.Martin Pieuchot
ok kettenis@, visa@
2018-01-04Unify <machine/mutex.h> a bit further.Martin Pieuchot
Remove `mtx_lock' from i386, add volatile before `mtx_owner' where it was missing. Inputs from kettenis@, ok visa@
2017-12-04Change __mp_lock_held() to work with an arbitrary CPU info structure andMartin Pieuchot
extend ddb(4) "ps /o" output to print which CPU is currently holding the KERNEL_LOCK(). Tested by dhill@, ok visa@
2017-09-05Move mutex, condvar, and thread-specific data routes, pthread_once, andPhilip Guenther
pthread_exit from libpthread to libc, along with low-level bits to support them. Major bump to both libc and libpthread. Requested by libressl team. Ports testing by naddy@ ok kettenis@
2017-04-20Hook up mutex(9) to witness(4).Visa Hankala
2017-04-13Provide mips64 with kernel-facing TCB_{GET,SET} macros that store itPhilip Guenther
in struct mdproc. With that, all archs have those and the __HAVE_MD_TCB macro can be unifdef'ed as always defined. ok kettenis@ visa@ jsing@
generated by cgit v1.2.3 (git 2.25.1) at 2024-11-29 02:56:18 +0000