TODO revision 1.46
1 /* $NetBSD: TODO,v 1.46 2022/01/05 16:01:54 andvar Exp $ */ 2 3 Things to be done: 4 5 common: 6 - make %g6 point to curcpu 7 - make %g7 point to curlwp 8 - change run-time checks for cpu type to function pointers 9 10 sun4u: 11 - GENERIC.UP kernel hangs on v445 (missing interrupt?) 12 13 sun4v: 14 - current status 15 T5 ldom with 2 VCPU and 4GB (primary ldom is Solaris 11.4 SRU30):: 16 - kernel boots from miniroot.fs via ldom disk (vdsk) 17 - ldom virtual network interface (vnet) is working 18 (verified by exiting sysinst and issuing a ping command) 19 - the sysinst tool starts, disk setup is working, network configuration is working, 20 but after the selected (minimal) sets have been downloaded and installed 21 the 'sh MAKEDEV all' command hangs. 22 T2000 ldom with 8 VCPU and 4GB: 23 - crashes in /sbin/init doing an access() call where %o0 is corrupted (zero) 24 S7 ldom with 8 VCPU and 16GB (primary ldom is Solaris 11.4 SRU33): 25 - same status as T5 ldom 26 - 64-bit kernel support 27 - 32-bit kernel support 28 - libkvm 29 - ofwboot: tlb_init_sun4v() hardcodes number of slots to 64 30 - locore.s: sun4v_datatrap missing implementation for trap level 1 31 - check build without SUN4V defined 32 - replace relevant references to %ver with GET_MAXCWP 33 - pmap_mp_init(): sun4v missing handling 34 - replace constructs like "wrpr %g0, PSTATE_KERN, %pstate" with NORMAL_GLOBALS 35 - replace constructs line "wrpr %g0, PSTATE_INTR, %pstate" with ALTERNATE_GOBALS 36 - sun4v tsb no need to lock... per cpu... anyway... 37 - ci_tsb_desc->td_ctxidx: -1 or 1? 38 - MP support - currently bypassed in pmap_bootstrap() for sun4v 39 - vpci.c/vpcivar.h: cleanup FIXMEs 40 - interrupts not handled properly (com at ebus only...) 41 - mpt(4) complains: mpt0: Phy 0: Link Status Unknown 42 - man pages for drivers imported from OpenBSD lke vpci, vbus, cbus, vdsk, ldc etc. 43 - vdsk and ldc drivers: code maked with OPENBSD_BUSDMA - make the bus_dma stuff work properly 44 - vbus.c: handle prom_getprop() memory leaks 45 - locore.s: rft_user (sun4v specific manual fill) - seems to work, but is it good enough (compared to openbsds rft_user? 46 - openfirmware interface 47 bounce handlng could need rework 48 - only #if __arch64__ independent for SUN4V, as OF for all ultrasparcs is 49 plain 32bit code too (and this #if only because the point below becomes 50 always false otherwise) 51 - only bounce if the address is >= 1UL<<32 52 - perhaps apply the lock to OF calls in general, not only the bounce'd ones 53 (in other words: let's protect the OF context with the lock, not 54 only the bounce buffer contents). 55 - Are there any other bounce-worthy OF calls? If so, apply handling 56 to all of them now 57 - consider method used by Solaris/OpenSolaris/Illumos (look for PROM_32BIT_ADDRS in 58 the Illumos codebase) 59 - platforms tested so far: 60 -- qemu sun4v/niagara - reaches the point where the kernel needs to mount a root fs 61 -- T2000 - mmu fault somewhere 62 -- T5 - stickcmpr is not properly initialized ??? 63
Indexes created Tue Sep 30 11:09:46 GMT 2025