.TITLE VQDRIVER - VAX/VMS VIRT DISK DRIVER w/shadowing .IDENT 'V1c' ; gce 6/22/1991 add some extra stuff to handle SMP full checking ; more carefully... ; .link 'sys$system:sys.stb'/sel ; ; Changed by making it fd1 type and adding nodename prefix characteristic ; so that hopefully it will now work with MSCP. ; ; Added save/restore of IRP$L_MEDIA for second I/O completion (by VDDRIVER) ; so that error path code that assumes this field unaltered will work. 4/14/89 ; ;md$stat=1 ;x$$$dt=1 ;call xdelta ; Note: define symbol VMS$V5 to assemble in VMS V5.x or later. Default ; assembly without this definition produces a VMS V4.x driver. ; define v5$picky also for normal SMP operations vms$v5=1 v5$picky=1 ; Glenn C. Everhart, 2/2/1989 ; Merged in some of Marty Sasaki's changes ;USAPADDR=0 ; ; FACILITY: ; ; VAX/VMS VIRTUAL DISK DRIVER USING CONTIGUOUS FILES. ; ; AUTHOR: ; ; G. EVERHART ; ; ; ABSTRACT: ; ; THIS MODULE CONTAINS THE TABLES AND ROUTINES NECESSARY TO ; PERFORM ALL DEVICE-DEPENDENT PROCESSING OF AN I/O REQUEST ; FOR VMS VIRTUAL DISKS ON CONTIG FILES. ; ; The VE: driver is intended to become a driver for software ; shadowed virtual disks. It will have several operation modes: ; 1. Only one file specified...works like VD: ; 2. Two files specified, startup mode...reads first file ; always, writes both files. (For the moment we leave ; it to the assign task to set disk size to the min ; of the two file sizes.) ; 3. Full shadow mode...reads one file or the other, depending ; on closest last LBN; writes all writes to both. ; In addition, to prevent errors during a "catchup" mode, a special ; r/w logical will be implemented. This will work exactly like ; normal r/w logical block most of the time, but when a special ; logical read is done, a flag will be set and the block number ; of the special read saved. When a special write is done, the ; flag will be cleared. ; IF an ordinary write to this block occurs (actually, to this ; 16 blocks...assume this granularity) so that the normal disk ; I/O may have interfered with the recovery, another flag will ; be set, and the special write will return an error. This can ; be used to repeat the read and try again, allowing normal ; access in mode 2 to the disk during catchup. Once a catchup ; has completed, the disk can be switched to mode 3 for better ; performance. Since a shadow disk SHOULD be on a different ; physical volume from the shadowed one, one HOPES that in ; general this will yield performance improvements. ; ; The IRP will come back to VDDRIVER twice for writes, being shipped ; off first to the first file's driver, then, after modification, ; to the second file's driver. After that it will have I/O completion ; run on it. Since the VE: unit stays busy during this whole time, ; flags in the UCB will be used to determine which of these two ; results are needed. In the case of reads, the start_io code will ; assign the IRP to one or the other file and the completion will ; just finish it up. ; ; ; The driver does not attempt to copy the IRP and issue the two ; I/O requests simultaneously; rather, it re-uses the IRP serially ; instead. Note for a striping driver, copying the IRP would be ; needed instead, and I/O completion would be done only when ; all parts had completed. The tricky part of copying the IRP is ; handling the IRPEs if any, by the way. To some extent FDDRV ; is a simpler engine on which to build a striping driver, though ; it will have lower performance than VDDRIVER could attain. ; ; Note: ; This driver has an FDT table that will look just like other ; disks for everything, but will NOT do buffered I/O. ; In its' FDT routines, which will be dummies, it will just ; modify logical block numbers in the I/O packets, assume ; the buffered bit in irp$w_STS clear (like all disk drivers ; except floppy drivers), do a range check to make sure ; the LBN used is in the legal range for this particular unit ; ov VD:, and reset things to call the real driver's FDT ; routines and let IT do the work. It will unbusy itself ; before losing control. ; The idea is that only I/O "gets at the physical ; storage", so only that need be munged. Since this happens ; only for read/write logical/physical, we just leave OUR ; FDT routines in there for everything else. In just assuming ; the buffered bit in the I/O we MAY mess up some quotas (this ; will eventually get cleaned up), but won't lose buffers or ; otherwise mess up things. Since most drivers have nonbuffered ; I/O, we'll usually be just fine. (Disk drivers, that is.) ; (In fact only RX01 class real drivers are buffered. Don't ; put vd: units on those!!!) ; To call the "real" FDT routines we'll just reset the ; registers (and the IORP) to the real device and return. The ; exec routines that handle FDT routines will take it from there ; and do the FDT processing in the real driver. We just have to ; let them work... ; The FDT routines are called from SYSQIOREQ.MAR for ; future reference. ; The main work is in the start-IO area. There, the I/O ; packet is edited to relocate the I/O to the contiguous file ; and sent to the real driver. First however we bash the IRP$L_PID ; field to get control back (via jsb) here. At that point, ; the UCB and IRP pointers are restored, the IRP returned to point ; at VD: (and the correct unit), and the normal IOC$REQCOM call ; is made to finish off the I/O packet in VD: context. (This allows ; any i/o splits to be handled correctly and SHOULD allow MSCP ; to work OK if enabled.) The vd: unit stays busy until reqcom ; calls free it. Normal reqcom completion in the host will free ; THAT unit also; we don't need to mess with it. The VD: unit will ; eventually RSB its way back to normal processing in the rest of ; VMS. ; ; Note: It is an interesting idea to have the delete FDT entry just ; return hardware error...a step toward a no-delete disk for keeping ; logfiles on. By checking a block was not written before (e.g. in ; an fddrv setting) one could make it a software write once medium ; too...one should do this above some "fence" block number so directories ; and bitmaps and index files could be updated. ; ; The logical I/O FDT entry has been commented out to keep the ; operation simpler. It is OK as is, but things are cleaner with ; NO prospect of clobbering any arguments. ; ; Note that ASNVD simply inserts a phony device structure for VD: units ; of 64 sectors per track, one track/cylinder, n cylinders. INIT seems ; to need this, though physical I/O is disabled. ;-- .PAGE .SBTTL EXTERNAL AND LOCAL DEFINITIONS ; ; EXTERNAL SYMBOLS ; .library /SYS$SHARE:LIB/ ; $ADPDEF ;DEFINE ADAPTER CONTROL BLOCK $CRBDEF ;DEFINE CHANNEL REQUEST BLOCK $DYNDEF ;define dynamic data types $DCDEF ;DEFINE DEVICE CLASS $DDBDEF ;DEFINE DEVICE DATA BLOCK $DEVDEF ;DEFINE DEVICE CHARACTERISTICS $DPTDEF ;DEFINE DRIVER PROLOGUE TABLE $EMBDEF ;DEFINE ERROR MESSAGE BUFFER $IDBDEF ;DEFINE INTERRUPT DATA BLOCK $IODEF ;DEFINE I/O FUNCTION CODES $DDTDEF ; DEFINE DISPATCH TBL... $ptedef $vadef $IRPDEF ;DEFINE I/O REQUEST PACKET $irpedef $PRDEF ;DEFINE PROCESSOR REGISTERS $SSDEF ;DEFINE SYSTEM STATUS CODES $UCBDEF ;DEFINE UNIT CONTROL BLOCK $VECDEF ;DEFINE INTERRUPT VECTOR BLOCK $pcbdef $jibdef p1=0 ; first qio param p2=4 p3=8 p4=12 p5=16 p6=20 ;6th qio param offsets .IF DF,VMS$V5 ;VMS V5 + LATER ONLY $SPLCODDEF $cpudef .ENDC ; ; UCB OFFSETS WHICH FOLLOW THE STANDARD UCB FIELDS ; $DEFINI UCB ;START OF UCB DEFINITIONS ;.=UCB$W_BCR+2 ;BEGIN DEFINITIONS AT END OF UCB .=UCB$K_LCL_DISK_LENGTH ;v4 def end of ucb ; USE THESE FIELDS TO HOLD OUR LOCAL DATA FOR VIRT DISK. $DEF UCB$W_vq_WPS .BLKW 1 ;Words per sector. $DEF UCB$W_vq_CS .BLKW 1 ;CONTROL STATUS REGISTER $DEF UCB$W_vq_DB .BLKW 1 ;UCB ADDRESS OF HOST DRIVER $DEF UCB$W_vq_DPN .BLKW 1 ;(LONGWORD) $DEF UCB$L_vq_DPR .BLKL 1 ;START LBN OF HOST CONTIG FILE $DEF UCB$L_vq_FMPR .BLKL 1 ; $DEF UCB$L_vq_PMPR .BLKL 1 ;PREVIOUS MAP REGISTER $DEF UCB$B_vq_ER .BLKB 1 ;SPECIAL ERROR REGISTER .BLKB 1 ;Reserved. $DEF UCB$B_vq_LCT .BLKB 1 ;LOOP COUNTER $DEF UCB$B_vq_XBA .BLKB 1 ;BUS ADDRESS EXTENSION BITS $DEF UCB$W_vq_PWC .BLKW 1 ;PARTIAL WORD COUNT $DEF UCB$W_vq_SBA .BLKW 1 ;SAVED BUFFER ADDRESS $DEF UCB$L_vq_XFER .BLKL 1 ;TRANSFER FUNCTION CSR BITS $DEF UCB$L_vq_LMEDIA .BLKL 1 ;LOGICAL MEDIA ADDRESS $DEF UCB$Q_vq_EXTENDED_STATUS ; Area into which we do READ ERROR .BLKQ 1 ; REGISTER command. ; use following field to hold irp$l_svapte and irp$l_bcnt ; across two transfers $DEF UCB$Q_vq_SVAPTETMP ; Area in which we save UCB fields - .BLKQ 1 ; SVAPTE, BOFF, and BCNT. $DEF UCB$L_vq_MAPREGTMP ; Area in which we save CRB fields - .BLKL 1 ; MAPREG, NUMREG, and DATAPATH. $DEF UCB$L_vq_SAVECS .BLKL 1 ; Area in which we save CS and DB regs. ; Add our stuff at the end to ensure we don't mess some fields up that some ; areas of VMS may want. $DEF UCB$HUCB .BLKL 1 ;ADDRESS OF HOST UCB $DEF UCB$HLBN .BLKL 1 ;LBN OF HOST FILE $DEF UCB$HFSZ .BLKL 1 ;SIZE OF HOST FILE, BLKS $DEF UCB$PPID .BLKL 1 ;PID OF ORIGINAL PROCESS FROM IRQ BLK ; NOTE: It is important to ENSURE that we never clobber IRP$L_PID twice! ; therefore, adopt convention that UCB$PPID is cleared whenever we put ; back the old PID value in the IRP. Only clobber the PID where ; UCB$PPID is zero!!! $DEF UCB$stats .BLKL 1 ;STATUS CODE SAVE AREA $DEF UCB$OBCT .BLKL 1 ;STORE FOR IRP$L_OBCNT too $def ucb$lmedia .blkl 1 ;storage for IRP$L_MEDIA $def ucb$owind .blkl 1 ; store irp$l_wind... $def ucb$osegv .blkl 1 ; and irp$l_segvbn ; Since I/O postprocessing on virtual or paging I/O makes lots of ; assumptions about location of window blocks, etc., which are ; not true here (wrong UCB mainly), we'll bash the function code ; we send to the host driver to look like logical I/O is being ; done and save the real function code here. Later when VD: does ; I/O completion processing, we'll replace the original function ; from here back in the IRP. This will be saved/restored along with ; ucb$ppid (irp$l_pid field) and so synchronization will be detected ; with ucb$ppid usage. ; define extra fork blks to try to avoid double forking possibilities $def ucb$l_vq_host_descr .blkl 2 ; char string descr ; $def ucb$vdcontfil .blkb 80 ; new structures for shadow disk operation $def ucb$hucb2 .blkl 1 ;host UCB of file 2 $def ucb$hlbn2 .blkl 1 ;LBN of file 2 $def ucb$hfsz2 .blkl 1 $def ucb$shmd .blkl 1 ;shadow mode. 0=use file 1 only ; 1= read file 1 always; write both ; 2= read both (select); write both $def ucb$rwlk .blkl 1 ; read/write interlock. Initialize to 1 ; on special read-logical, set to 44 on ; normal write in area of block read, ; return as i/o status and clear on ; special write-logical. $def ucb$rwlbn .blkl 1 ; LBN of first block in special read/logical $def ucb$rwsz .blkl 1 ; size of special transfer $def ucb$llbn1 .blkl 1 ; last LBN of file 1 $def ucb$llbn2 .blkl 1 ; last LBN of file 2 $def ucb$ucbos .blkl 1 ; offset to ucb/lbn/fsz in ucb to use at fin-io $def ucb$rwdir .blkl 1 ;read (0) or write (1) I/O $def ucb$er1 .blkl 1 ; store for error code of 1st write of 2 ; ;end new structures for shadowing ; $DEF UCB$K_vq_LEN .BLKW 1 ;LENGTH OF UCB ;UCB$K_vq_LEN=. ;LENGTH OF UCB $DEFEND UCB ;END OF UCB DEFINITONS .SBTTL STANDARD TABLES ; ; DRIVER PROLOGUE TABLE ; ; THE DPT DESCRIBES DRIVER PARAMETERS AND I/O DATABASE FIELDS ; THAT ARE TO BE INITIALIZED DURING DRIVER LOADING AND RELOADING ; .PSECT $$$105_PROLOGUE vq_UNITS=8 VE$DPT:: DPTAB - ;DPT CREATION MACRO END=vq_END,- ;END OF DRIVER LABEL ADAPTER=NULL,- ;ADAPTER TYPE = NONE (VIRTUAL) FLAGS=DPT$M_SMPMOD,- ;smp ok DEFUNITS=2,- ;UNITS 0 THRU 1 UCBSIZE=UCB$K_vq_LEN,- ;LENGTH OF UCB MAXUNITS=vq_UNITS,- ;FOR SANITY...CAN CHANGE NAME=VQDRIVER ;DRIVER NAME DPT_STORE INIT ;START CONTROL BLOCK INIT VALUES DPT_STORE DDB,DDB$L_ACPD,L,<^A\F11\> ;DEFAULT ACP NAME DPT_STORE DDB,DDB$L_ACPD+3,B,DDB$K_PACK ;ACP CLASS .IF NDF,VMS$V5 DPT_STORE UCB,UCB$B_FIPL,B,8 ;FORK IPL (VMS V4.X) .IFF ;DEFINE FOR VMS V5.X & LATER DPT_STORE UCB,UCB$B_FLCK,B,SPL$C_IOLOCK8 ;FORK IPL (VMS V5.X + LATER) .ENDC ; NOTE THESE CHARACTERISTICS HAVE TO LOOK LIKE THE "REAL" DISK. DPT_STORE UCB,UCB$L_DEVCHAR,L,- ;DEVICE CHARACTERISTICS ; RANDOM ACCESS DPT_STORE UCB,UCB$L_DEVCHAR2,L,- ;DEVICE CHARACTERISTICS ; Prefix name with "node$" (like rp06) DPT_STORE UCB,UCB$B_DEVCLASS,B,DC$_DISK ;DEVICE CLASS DPT_STORE UCB,UCB$W_DEVBUFSIZ,W,512 ;DEFAULT BUFFER SIZE ; FOLLOWING DEFINES OUR DEVICE "PHYSICAL LAYOUT". It's faked here and ; this structure (64 sectors/trk, 1 trk/cyl, nn cylinders) forces ; VE: units to be in multiples of 64 blocks. It can be modified as ; appropriate. However, recall that one has 1 byte for sectors/trk ; and 16 bits for cylinder number and 1 byte for tracks/cylinder. ; The current structure allows vd: units as large as 65535*64 blocks ; (about 4 million blocks, or 2 gigabytes), which is probably big enough ; for most purposes. The actual size is set up in ASNVD which finds the ; number of cylinders to "fit" in the container file. For emulating other ; ODS-2 volumes, the appropriate physical structure should be emulated also. ; NO logic in this driver depends on this stuff. It just has to be there ; to keep INIT and friends happy. ; (In fact the current ASNVD image for VD: does this. We however ; have to preset SOMETHING.) DPT_STORE UCB,UCB$B_TRACKS,B,1 ; 1 TRK/CYL DPT_STORE UCB,UCB$B_SECTORS,B,64 ;NUMBER OF SECTORS PER TRACK DPT_STORE UCB,UCB$W_CYLINDERS,W,16 ;NUMBER OF CYLINDERS ; FAKE GEOMETRY TO MAKE TRANSLATION EASIER. HAVE PRIV'D IMAGE LATER ; RESET THE UCB$W_CYLINDERS TO WHATEVER'S DESIRED. JUST MAKE SURE IT'S ; A MULTIPLE OF 64 BLOCKS IN SIZE, WHICH OUGHT TO BE GOOD ENOUGH. DPT_STORE UCB,UCB$B_DIPL,B,21 ;DEVICE IPL DPT_STORE UCB,UCB$B_ERTMAX,B,10 ;MAX ERROR RETRY COUNT DPT_STORE UCB,UCB$W_DEVSTS,W,- ;INHIBIT LOG TO PHYS CONVERSION IN FDT ;... ; ; don't mess with LBN; leave alone so it's easier to hack on... ; DPT_STORE REINIT ;START CONTROL BLOCK RE-INIT VALUES ; DPT_STORE CRB,CRB$L_INTD+VEC$L_ISR,D,vq_INT ;INTERRUPT SERVICE ROUTINE ADDRESS DPT_STORE CRB,CRB$L_INTD+VEC$L_INITIAL,- ;CONTROLLER INIT ADDRESS D,vq_ctrl_INIT ;... DPT_STORE CRB,CRB$L_INTD+VEC$L_UNITINIT,- ;UNIT INIT ADDRESS D,vq_unit_INIT ;... DPT_STORE DDB,DDB$L_DDT,D,VE$DDT ;DDT ADDRESS DPT_STORE END ;END OF INITIALIZATION TABLE ; ; DRIVER DISPATCH TABLE ; ; THE DDT LISTS ENTRY POINTS FOR DRIVER SUBROUTINES WHICH ARE ; CALLED BY THE OPERATING SYSTEM. ; ;VD$DDT: DDTAB - ;DDT CREATION MACRO DEVNAM=VE,- ;NAME OF DEVICE START=vq_STARTIO,- ;START I/O ROUTINE FUNCTB=vq_FUNCTABLE,- ;FUNCTION DECISION TABLE ; CANCEL=0,- ;CANCEL=NO-OP FOR FILES DEVICE ; REGDMP=0,- ;REGISTER DUMP ROUTINE ; DIAGBF=0,- ;BYTES IN DIAG BUFFER ERLGBF=0 ;BYTES IN ;ERRLOG BUFFER ; ; FUNCTION DECISION TABLE ; ; THE FDT LISTS VALID FUNCTION CODES, SPECIFIES WHICH ; CODES ARE BUFFERED, AND DESIGNATES SUBROUTINES TO ; PERFORM PREPROCESSING FOR PARTICULAR FUNCTIONS. ; vq_FUNCTABLE: FUNCTAB ,- ;LIST LEGAL FUNCTIONS ; MOUNT VOLUME ; no-op phys I/O for a test here... FUNCTAB ,- ;BUFFERED FUNCTIONS ; MOUNT VOLUME FUNCTAB vq_ALIGN,- ;TEST ALIGNMENT FUNCTIONS functab vq_format,- ;point to host disk ; NOTE SEPARATE CALL FOR PHYSICAL I/O SO WE CAN JUST CONVERT TO LOGICAL AND ; DO OUR THING... CONVERT TO A LOGICAL QIO THERE FOR "REAL" DRIVER ALSO ; SO IT CAN DO CONVERSION TO ITS IDEA OF PHYSICAL IF IT WISHES... ; ; LEAVE NORMAL ACP CALLS IN SO FILE STRUCTURED STUFF ON OUR VE: UNIT ; WILL WORK OK. ; Functab vq_rdspc,- ;read - special logic here. Functab vq_wrtspc,- ;write - special logic here. FUNCTAB +ACP$READBLK,- ;READ FUNCTIONS FUNCTAB +ACP$WRITEBLK,- ;WRITE FUNCTIONS FUNCTAB +ACP$ACCESS,- ;ACCESS FUNCTIONS FUNCTAB +ACP$DEACCESS,- ;DEACCESS FUNCTION FUNCTAB +ACP$MODIFY,- ;MODIFY FUNCTIONS FUNCTAB +ACP$MOUNT,- ;MOUNT FUNCTION ; MOUNT VOLUME FUNCTAB +EXE$ZEROPARM,- ;ZERO PARAMETER FUNCTIONS ; AVAILABLE FUNCTAB +EXE$ONEPARM,- ;ONE PARAMETER FUNCTION FUNCTAB +EXE$SENSEMODE,- ;SENSE FUNCTIONS FUNCTAB +EXE$SETCHAR,- ;SET FUNCTIONS .PAGE .SBTTL FDT Routines ;++ ; ; vq_format - point to proper location on the host disk ; ; With no function modifiers, this routine takes as arguments the name ; of the host disk (the real disk where the virtual disk will exist), ; the size of the virtual disk, and the LBN where the virtual disk ; will start. After these are set up, the device is put online and is ; software enabled. ; ; This routine does virtually no checking, so the parameters must be ; correct. ; ; Inputs: ; p1 - pointer to buffer. The buffer has the following format: ; longword 0 - starting LBN, where the virtual disk starts ; on the real disk. ; longword 1 - virtual disk length, the number of blocks in ; the virtual disk. ; longword 2 through the end of the buffer, the name of the ; virtual disk. This buffer must be blank ; padded if padding is necessary ; ; p2 - size of the above buffer ;-- vq_format: bicw3 #io$m_fcode,irp$w_func(r3),r0 ;mask off function code bneq 20$ ;branch if modifiers, special ;thus, normal io$_format will do nothing. rsb ;regular processing 10$: movzwl #SS$_BADPARAM,r0 ;illegal parameter clrl r1 jmp g^exe$abortio 20$: movl p1(ap),r0 ;buffer address movl p2(ap),r1 ;length of buffer jsb g^exe$writechk ;read access? doesn't return on error clrl irp$l_bcnt ;paranoia, don't need to do this... movl p1(ap),r0 ;get buffer address movl (r0)+,- ;move starting lbn ucb$hlbn(r5) blss 10$ movl (r0)+,- ;size of virtual disk ucb$hfsz(r5) bleq 10$ movl (r0),- ;name of "real" disk ucb$l_vq_host_descr+4(r5) subl3 #8,p2(ap),- ;set length of name in descriptor ucb$l_vq_host_descr(r5) bleq 10$ ;bad length moval ucb$l_vq_host_descr(r5),r1 ;descriptor for... jsb g^ioc$searchdev ;search for host device blbs r0,30$ ;branch on success movzwl #ss$_nosuchdev+2,r0 ;make an error, usually a warning clrl r1 jmp g^exe$abortio ;exit with error 30$: addl3 ucb$hfsz(r5),- ;end of virtual device ucb$hlbn(r5),r0 cmpl ucb$l_maxblock(r1),r0 ; < end of real disk? blss 10$ movl r1,ucb$hucb(r5) ;stash the ucb clrl ucb$ppid(r5) ; mark driver free of old pids bisw #ucb$m_valid,ucb$w_sts(r5) ;set volume valid bisw #ucb$m_online,ucb$w_sts(r5) ;set unit online movl ucb$l_irp(r5),r3 ;restore r3, neatness counts movzwl #ss$_normal,r0 ;success jmp g^exe$finishioc ;wrap things up. ; Read-special logic vq_rdspc: bitw #128,irp$w_func(r3) ;mask off function code bneq 20$ ;branch if modifiers, special ; normal read-logical just returns rsb ;regular processing 20$: movl #1,ucb$rwlk(r5) ;set up special interlock movl irp$l_media(r3),ucb$rwlbn(r5) ;store LBN we're writing at (start) movl irp$l_bcnt(r3),r0 ;get transfer size in bytes addl2 #511,r0 ;add blksize - 1 byte ashl #-9,r0,r0 ;get number blks in transfer movl r0,ucb$rwsz(r5) ;store size in blocks of transfer rsb ; and exit... ; ; write-special logic vq_wrtspc: ; check only 128 and 256 bits. bitw #128,irp$w_func(r3) ;mask off function code bneq 20$ ;branch if modifiers, special ; This is a NORMAL write-logical-block. Therefore, check if it overlaps ; a special read area just handled. jsb ovlpchk ; do the overlap check ; ucb$rwlk returns 44 if fails. ; assumes r3 -> IRP, r5 -> UCB ; normal write-logical just returns rsb ;regular processing 20$: blbc ucb$rwlk(r5),30$ ; on error, abort the I/O clrl ucb$rwlk(r5) ; reset the special mode flag rsb ; and exit... 30$: ; normal I/O interfered with the ; interlocked r/w. Return an error. ; pick random error 44 as our signal that the I/O failed. clrl ucb$rwlk(r5) ; reset the special mode flag movzwl #44,r0 ;make an error. Catchup prog needs to re-read. clrl r1 jmp g^exe$abortio ;exit with error ; Note we have a back-up check of this stuff in the start-io entry area ; in case anything slips by due to nonzero queue length for the ve: ; driver. ; ovlpchk - check for overlap of write with what was read ; assumes r5 is UCB address, r3 is IRP address and that it is ; a write-logical IRP. ovlpchk: tstl ucb$rwlk(r5) ; special read done? beql 99$ ; if not, just return. pushl r0 pushl r1 pushl r2 pushl r3 ; need a few regs movl irp$l_media(r3),r1 ; LBN we're writing at (start) movl irp$l_bcnt(r3),r0 ;get transfer size in bytes addl2 #511,r0 ;add blksize - 1 byte ashl #-9,r0,r0 ;get number blks in transfer ; r0 now is # blks we have addl2 r1,r0 ; r0 is now end LBN of this request ; r1 = start LBN of request, r0 = end LBN movl ucb$rwlbn(r5),r2 ; low LBN of read-special addl3 ucb$rwsz(r5),r2,r3 ; r3 = high LBN of read-special ; if our high LBN is below the low LBN of the r-s we don't overlap. ; If our low LBN is above the high LBN of the r-s we don't overlap. ; Otherwise we overlap. cmpl r0,r2 ;our high less than r-s low? blssu 99$ ; if so nothing to do cmpl r1,r3 ;our low greater than r-s high? bgtru 99$ ; if so nothing to do movl #44,ucb$rwlk(r5) ;if OVERLAP then flag error for w-s next/ popl r3 popl r2 popl r1 popl r0 99$: rsb .SBTTL CONTROLLER INITIALIZATION ROUTINE ; ++ ; ; vq_ctrl_INIT - CONTROLLER INITIALIZATION ROUTINE ; ; FUNCTIONAL DESCRIPTION: ; noop ; INPUTS: ; R4 - CSR ADDRESS ; R5 - IDB ADDRESS ; R6 - DDB ADDRESS ; R8 - CRB ADDRESS ; ; THE OPERATING SYSTEM CALLS THIS ROUTINE: ; - AT SYSTEM STARTUP ; - DURING DRIVER LOADING ; - DURING RECOVERY FROM POWER FAILURE ; THE DRIVER CALLS THIS ROUTINE TO INIT AFTER AN NXM ERROR. ;-- .PSECT $$$115_DRIVER vq_ctrl_INIT: ;vd CONTROLLER INITIALIZATION CLRL CRB$L_AUXSTRUC(R8) ; SAY NO AUX MEM RSB ;RETURN .PAGE .SBTTL INTERNAL CONTROLLER RE-INITIALIZATION ; ; INPUTS: ; R4 => controller CSR (dummy) ; R5 => UCB ; ctrl_REINIT: RSB ; RETURN TO CALLER .PAGE .SBTTL UNIT INITIALIZATION ROUTINE ;++ ; ; vq_unit_INIT - UNIT INITIALIZATION ROUTINE ; ; FUNCTIONAL DESCRIPTION: ; ; THIS ROUTINE SETS THE VD: ONLINE. ; ; THE OPERATING SYSTEM CALLS THIS ROUTINE: ; - AT SYSTEM STARTUP ; - DURING DRIVER LOADING ; - DURING RECOVERY FROM POWER FAILURE ; ; INPUTS: ; ; R4 - CSR ADDRESS (CONTROLLER STATUS REGISTER) ; R5 - UCB ADDRESS (UNIT CONTROL BLOCK) ; R8 - CRB ADDRESS ; ; OUTPUTS: ; ; THE UNIT IS SET ONLINE. ; ALL GENERAL REGISTERS (R0-R15) ARE PRESERVED. ; ;-- vq_unit_INIT: ;vd UNIT INITIALIZATION ; Don't set unit online here. Priv'd task that assigns VD unit ; to a file does this to ensure only assigned VDn: get used. ; BISW #UCB$M_ONLINE,UCB$W_STS(R5) ;SET UCB STATUS ONLINE ;limit size of VD: data buffers vq_bufsiz=8192 ; (asnvq will set this to same as host...) movl #vq_bufsiz,ucb$l_maxbcnt(r5) ;limit transfers to 8k MOVB #DC$_DISK,UCB$B_DEVCLASS(R5) ;SET DISK DEVICE CLASS ; NOTE: we may want to set this as something other than an RX class ; disk if MSCP is to use it. MSCP explicitly will NOT serve an ; RX type device. For now leave it in, but others can alter. ; (There's no GOOD reason to disable MSCP, but care!!!) movl #^Xb12c5001,ucb$l_media_id(r5) ; set media id as VE ; (note the id might be wrong but is attempt to get it.) (used only for ; MSCP serving.) MOVB #DT$_FD1,UCB$B_DEVTYPE(R5) ;Make it foreign disk type 1 ; (dt$_rp06 works but may confuse analyze/disk) ;;; NOTE: changed from fd1 type so MSCP will know it's a local disk and ;;; attempt no weird jiggery-pokery with the VD: device. ; MSCP may still refuse to do a foreign drive too; jiggery-pokery later ; to test if there's occasion to do so. RSB ;RETURN .PAGE .SBTTL FDT ROUTINES ;++ ; ; vq_ALIGN - FDT ROUTINE TO TEST XFER BYTE COUNT ; ; FUNCTIONAL DESCRIPTION: ; ; THIS ROUTINE IS CALLED FROM THE FUNCTION DECISION TABLE DISPATCHER ; TO CHECK THE BYTE COUNT PARAMETER SPECIFIED BY THE USER PROCESS ; FOR AN EVEN NUMBER OF BYTES (WORD BOUNDARY). ; ; INPUTS: ; ; R3 - IRP ADDRESS (I/O REQUEST PACKET) ; R4 - PCB ADDRESS (PROCESS CONTROL BLOCK) ; R5 - UCB ADDRESS (UNIT CONTROL BLOCK) ; R6 - CCB ADDRESS (CHANNEL CONTROL BLOCK) ; R7 - BIT NUMBER OF THE I/O FUNCTION CODE ; R8 - ADDRESS OF FDT TABLE ENTRY FOR THIS ROUTINE ; 4(AP) - ADDRESS OF FIRST FUNCTION DEPENDENT QIO PARAMETER ; ; OUTPUTS: ; ; IF THE QIO BYTE COUNT PARAMETER IS ODD, THE I/O OPERATION IS ; TERMINATED WITH AN ERROR. IF IT IS EVEN, CONTROL IS RETURNED ; TO THE FDT DISPATCHER. ; ;-- nolchk=1 vq_ALIGN: ;CHECK BYTE COUNT AT P1(AP) .if ndf,nolchk ; note: not fully tested but a MINOR mod... therefore conditioned. ; tstw 6(ap) ;test high order half of ; byte count specified ; bneq 10$ ; if bigger than 65k call error .endc BLBS 4(AP),10$ ;IF LBS - ODD BYTE COUNT RSB ;EVEN - RETURN TO CALLER 10$: MOVZWL #SS$_IVBUFLEN,R0 ;SET BUFFER ALIGNMENT STATUS JMP G^EXE$ABORTIO ;ABORT I/O .PAGE .SBTTL START I/O ROUTINE ;++ ; ; vq_STARTIO - START I/O ROUTINE ; ; FUNCTIONAL DESCRIPTION: ; ; THIS FORK PROCESS IS ENTERED FROM THE EXECUTIVE AFTER AN I/O REQUEST ; PACKET HAS BEEN DEQUEUED. ; ; INPUTS: ; ; R3 - IRP ADDRESS (I/O REQUEST PACKET) ; R5 - UCB ADDRESS (UNIT CONTROL BLOCK) ; IRP$L_MEDIA - PARAMETER LONGWORD (LOGICAL BLOCK NUMBER) ; ; OUTPUTS: ; ; R0 - FIRST I/O STATUS LONGWORD: STATUS CODE & BYTES XFERED ; R1 - SECOND I/O STATUS LONGWORD: 0 FOR DISKS ; ; THE I/O FUNCTION IS EXECUTED. ; ; ALL REGISTERS EXCEPT R0-R4 ARE PRESERVED. ; ;-- .if df,x$$$dt rwflg: .long 0 .endc ; In fact, using XDelta, I have *never* seen requeue called. REQUEUE: .if df,x$$$dt jsb g^ini$brk .endc JMP G^EXE$INSIOQ ; REQUEUE packet to ourselves ; return to our caller direct from insioq. ; (note this also sets busy, so it will NOT loop forever.) VQ_STARTIO: ;START I/O OPERATION ; ; PREPROCESS UCB FIELDS ; ; ASSUME RY_EXTENDED_STATUS_LENGTH EQ 8 ; CLRQ UCB$Q_vq_EXTENDED_STATUS(R5) ; Zero READ ERROR REGISTER area. ; ; BRANCH TO FUNCTION EXECUTION bbs #ucb$v_online,- ; if online set software valid ucb$w_sts(r5),210$ 216$: movzwl #ss$_volinv,r0 ; else set volume invalid brw resetxfr ; reset byte count & exit 210$: tstl ucb$hucb(r5) ; do we have any host device? beql 216$ ; if eql no, flag invalid volume. ; THIS IS SAFETY FROM CONFIGURING FROM OUTSIDE ; BEFORE GOING ON, WE WANT TO ENSURE THE UCB IS FREE. TSTL UCB$PPID(R5) ; MAKE SURE we haven't got ; a packet in process BNEQ REQUEUE ; IF a packet's in process, requeue ; back to this driver; do NOT process ; immediately! ; Note...never seems to get to requeue (xdelta would catch it!) ; (that's a good sign; should never get there.) bisw #ucb$m_online,ucb$w_sts(r5) ; set online bisw #ucb$m_valid,ucb$w_sts(r5) ;set valid ; set ourselves as owners of channel for VD: movl ucb$l_crb(r5),r0 movl crb$l_intd+vec$l_idb(r0),r0 ;get idb address cmpl r5,idb$l_owner(r0) ;are we owners? beql 214$ ; if eql yes, all's well ; REQPCHAN ; gain access to controller in "standard" way 214$: ; 10$:; BBS #IRP$V_PHYSIO,- ;IF SET - PHYSICAL I/O FUNCTION ; IRP$W_STS(R3),20$ ;... BBS #UCB$V_VALID,- ;IF SET - VOLUME SOFTWARE VALID UCB$W_STS(R5),20$ ;... MOVZWL #SS$_VOLINV,R0 ;SET VOLUME INVALID STATUS BRW RESETXFR ;RESET BYTE COUNT AND EXIT 20$: ; IF WE GET A SEGMENT TRANSFER HERE (LOGICAL I/O) ; IT MUST BE UPDATED FOR HOST AND SHIPPED OUT. ; OUR UCB HAS BLOCK NUMBER INFO... ; FIND OUT IF THIS IS LOGICAL OR PHYSICAL I/O FIRST. THEN IF IT IS BUGGER ; THE I/O PACKET USING UCB INFO AND SEND TO THE REAL DRIVER... ; ALSO ENSURE WE ARE UNBUSIED... ; ; Sanity check for assignments: hucb2 field zero implies shmd must be ; zero. tstl ucb$hucb2(r5) ;second UCB there? bneq 21$ ;if neq yes clrl ucb$shmd(r5) ;if not, ensure in single-file mode 21$: ; EXTZV #IRP$V_FCODE,#IRP$S_FCODE,IRP$W_FUNC(R3),R1 ; GET FCN CODE case r1,<- ; Dispatch to function handling routine unload,- ; Unload nop,- ; Seek NOP,- ; Recalibrate(unsupported) nop,- ; Drive clear NOP,- ; Release port(unsupported) NOP,- ; Offset heads(unsupported) NOP,- ; Return to center nop,- ; Pack acknowledge NOP,- ; Search(unsupported) NOP,- ; Write check(unsupported) WRITEDATA,- ; Write data READDATA,- ; Read data NOP,- ; Write header(unsupported) NOP,- ; Read header(unsupported) NOP,- ; Place holder NOP,- ; Place holder available,- ; Available (17) NOP,NOP,NOP,- ; 18-20 NOP,NOP,NOP,NOP,nop,nop,nop,NOP,NOP,nop,- ;21-30 NOP,NOP,NOP,NOP,nop,NOP,nop,nop,nop,NOP,- ;31-40 NOP,NOP,NOP,NOP,NOP,NOP,NOP,NOP,NOP,nop,- ;41-50 NOP,NOP,NOP,NOP,nop,NOP,NOP,NOP,NOP,NOP,- ;51-60 nop,- ;61 >,LIMIT=#1 nop: ;unimplemented function brw fexl rwcmnj: brw rwcmn readdata: movl #ucb$hucb,ucb$ucbos(r5) ;store address of file 1 clrl ucb$rwdir(r5) ;set read cmpl ucb$shmd(r5),#1 ;in single disk mode or single read mode? bleq rwcmnj ;find which file to use for read. ; set up ucbos field to offset for that file. pushl r0 pushl r1 movl irp$l_media(r3),r0 ;get proposed lbn movl r0,r1 subl2 ucb$llbn1(r5),r0 ;get diffs of (new LBN) - (last LBNi) subl2 ucb$llbn2(r5),r1 ;now get abs values tstl r0 bgeq 1$ ;if positive leave alone mnegl r0,r0 1$: tstl r1 bgeq 2$ mnegl r1,r1 ;negate if neg. 2$: ; now r0 = "distance" on file 1, r1 = "distance" on file 2 cmpl r0,r1 ;which is greater? blssu 3$ ;if file 1 is less, leave file 1 in ucbos field movl #ucb$hucb2,ucb$ucbos(r5) ;if file 2 less, set that. movl irp$l_media(r3),ucb$llbn2(r5) ;store last LBN as this one, file 2 brb 4$ 3$: movl irp$l_media(r3),ucb$llbn1(r5) ;store last LBN as this one, file 1 4$: popl r1 popl r0 brw rwcmn writedata: .if df,x$$$dt jsb g^ini$brk .endc movl #ucb$hucb,ucb$ucbos(r5) ;store address of file 1 movl #1,ucb$rwdir(r5) ;set write movl irp$l_media(r3),ucb$llbn1(r5) ;store current as last LBN of both movl irp$l_media(r3),ucb$llbn2(r5) ;files for write since both get used ; Now one of the key parts: ; Check to see if this write is a NORMAL access that overlaps data set ; in rwlbn/rwsz. If so, set rwlk field to 44 so we can flag that a write ; has interfered with the data being caught-up. ; To avoid possible other holes, an FDT time check will also exist so that a ; user write that would invalidate the special write will be caught at FDT ; time also. bitw #128,irp$w_func(r3) ;mask off function code ; Check only 128 and 256 bits for function modifiers. ; Asnve and Asnvq use 128 bit only. ;this a normal write-logical? bneq rwcmn2 ;if modified, no overlap chk ;since this IS our write-special. jsb ovlpchk ; do the overlap check brw rwcmn rwcmn2: ; look over write-special HERE also to ensure that if we (somehow) get ; by the FDT time check, we don't let this by if we've had an overlapped ; normal write. ; (It suffices to check the ucb$rwlk field to see if it's low bit is clear) tstl ucb$rwlk(r5) ;special read done? beql rwcmn4 ;if eq no, no check needed cmpl ucb$rwlk(r5),#44 ;error flagged? bneq RWCMN4 ;if not, go ahead and finish it ; oops... ; We flagged an error. DON'T let this write go through! MOVZWL #44,R0 ;SET VOLUME INVALID STATUS clrl ucb$rwlk(r5) ;clear special field for next time BRW RESETXFR ;RESET BYTE COUNT AND EXIT rwcmn4: ; I/o will go through. However, this is a SHADOW write, so there isn't ; really any need to write the first disk over again. Just write the ;second... .if df,shdopt ; Just did FIRST of two writes. Fix things up and set up for the ; SECOND write. movl #ucb$hucb2,ucb$ucbos(r5) ;set up for second pass ; Issue of what to do with errors arises here. Save first host's ; driver status in ucb$er1 field here. movl #1,ucb$er1(r5) ;store original error as success .endc rwcmn: ; debug using sda to peek ; NOW VALIDATED I/O FCN... MODIFY AND SEND OFF CMPL IRP$L_MEDIA(R3),UCB$HFSZ(R5) ;BE SURE LBN OK blequ 65$ brw Fatalerr ;dismiss I/o if not ok block number 65$: ; HAVE TO BE CAREFUL WHAT WE SHIP TO REAL DRIVER ; Now that we know IRP$L_MEDIA is ok in IRP, save it for restore at ; I/O completion by VDDRIVER ; (This also facilitates remodification for second pass on writes) movl irp$l_media(r3),ucb$lmedia(r5) ; Prepare to enter another context. ; SEND PKT OFF TO REAL DRIVER... ; First get address stuff to bash PID with MOVL IRP$L_PID(R3),UCB$PPID(R5) ; SAVE PROCESS ID IN VD: UCB movzwl irp$w_sts(r3),ucb$stats(r5) ;save original fcn code movl irp$l_obcnt(r3),ucb$obct(r5) ;store obcnt field movl irp$l_svapte(r3),ucb$q_vq_svaptetmp(r5) ;save svapte movl irp$l_bcnt(r3),ucb$q_vq_svaptetmp+4(r5) ;and orig bcnt ; (actually shouldn't need to worry about driver post processing... ; we do all that here via hack to change irp$l_pid field....) ; movl irp$l_bcnt(r3),irp$l_obcnt(r3) ;and reset to actual ; requested so driver NEVER sees ; need to do postprocessing requeues ; in host context. (we do that in OUR ; context.) movl irp$l_wind(r3),ucb$owind(r5) ;store window ptr movl irp$l_segvbn(r3),ucb$osegv(r5) ;store segment vbn also MOVZWL UCB$W_UNIT(R5),-(SP) ; BUILD ADDRESS OF UCB STORE ASHL #2,(SP),(SP) ; WITH OFFSET * 4 MOVAB VQ_UCBTBL,-(SP) ; GET TBL BASE IN STACK ADDL2 (SP)+,(SP) ; NOW ADD BASE + OFFSET MOVL R5,@(SP)+ ; AND STORE UCB ADDRESS IN VQ_UCBTBL ; (THIS ALLOWS US TO GET IT BACK...) MOVZWL UCB$W_UNIT(R5),-(SP) ; BUILD ADDRESS OF ENTRY NOW MULL2 #vq_FXPL,(SP) ; MULTIPLY OFFSET BY SIZE OF ENTRY MOVAB vq_FXS0,IRP$L_PID(R3) ;AND BASH PID IN IRP SO WE ; GET BACK CONTROL AT vq_FIXSPLIT (VIA JSB) ; WHEN HOST'S I/O IS DONE. ADDL2 (SP)+,IRP$L_PID(R3) ;SET TO ENTER IN CORRECT ; UNIT'S ENTRY cmpl #ucb$hucb,ucb$ucbos(r5) ; bneq 68$ ; if using file 2, branch. ADDL2 UCB$HLBN(R5),IRP$L_MEDIA(R3) ;ADJUST LBN IN IO PKT ; ; NOW we have to fix up the media address for the host... ; ... otherwise we confuse the heck out of things by making ; ... the host (who is expecting a track/sect/cyl number) get really ; ... goofy numbers. Cheat by using exec routine after a bit more messup. ; Ideally we should also avoid moving the packet (next) to host if ; any segmented I/O may occur but ignore that for now; it will be ; rare and I want to get this working in the 99.9% cases... MOVL UCB$HUCB(R5),IRP$L_UCB(R3) ;FIX UP PTR IN I/O PKT ; GRAB HOST PID .if df,vms$v5 ;smp stuff .if df,relacq ;normally don't define this movzbl ucb$b_flck(r5),-(sp) ;save lock id .endc .endc MOVL UCB$HUCB(R5),R5 ;NOW POINT AT HOST UCB OURSELVES brb 69$ 68$: ;note: ONLY get here when we're accessing file 2 first for read. ADDL2 UCB$HLBN2(R5),IRP$L_MEDIA(R3) ;ADJUST LBN IN IO PKT ; MOVL UCB$HUCB2(R5),IRP$L_UCB(R3) ;FIX UP PTR IN I/O PKT .if df,vms$v5 ;smp stuff .if df,relacq ;normally don't define relacq movzbl fkb$b_flck(r5),-(sp) ;save lock id .endc .endc MOVL UCB$HUCB2(R5),R5 ;NOW POINT AT HOST UCB OURSELVES 69$: MOVL IRP$L_MEDIA(R3),R0 ;GET LBN TO CONVERT ; Note that the host driver normally will get physical I/O addresses ; in this entry. Logical I/O is converted to physical in FDT ; routines for most drivers; the few exceptions inhibit conversion ; via IOC$CVTLOGPHY anyway. Therefore we must ALWAYS convert to ; physical. JSB G^IOC$CVTLOGPHY ; LET THE EXEC DO IT ; Logical I/O... relocate it here ; Already adjusted the logical blk # earlier ; next op may mess up some regs. Also we cannot access the packet once ; we give it to the host driver thus: .if ndf,vms$v5 ;ok for v4 and non smp v5 JMP G^EXE$INSIOQ ; INSERT PACKET INTO HOST'S QUEUE .iff .if df,relacq .if ndf,v5$picky movl (sp)+,r0 ; get back lock ID ; note we stay busy here in vq unit so are protected against re-entry ; even though we drop briefly to IPL 0. forkunlock R0 ; release VQ: lock JMP G^EXE$INSIOQ ; INSERT PACKET INTO HOST'S QUEUE .iff movl (sp),r0 ;picky mode--- unlock vd: database here forkunlock r0 jsb g^exe$insioq ;put IRP on host's queue ;now the picky part...re-acquire VD: forklock before return to ; OUR exe$insioq can release it movl (sp)+,r0 ;our fork index forklock r0 ;go get lock again rsb ;now back to our caller for cleanup .endc ;v5$picky .iff ;relacq JMP G^EXE$INSIOQC ; insert packet and keep forklock ;this reliably solves SMP problems .endc ;relacq .endc ; WE Now have queued the work to the real driver. Since the ; I/O may have splits, just await done return and let the ; vq_fixsplit processing get done our cleanup. Because we need ; to await this, just return with VD: unit STILL BUSY to ensure ; that we don't get thru here until we're GOOD AND READY! 402$: BRW FEXL ;Else, branch to execute function. ; ; UNLOAD and AVAILABLE Functions ; Clear UCB$V_VALID in UCB$W_STS ; UNLOAD: AVAILABLE: ; BICW #UCB$M_VALID, - ;Clear sofware volume valid bit. ; UCB$W_STS(R5) ; BRB NORMAL ;Then complete the operation. ; ; OPERATON COMPLETION ; FEXL: ; dummy entry ... should never get here NORMAL: ;SUCCESSFUL OPERATION COMPLETE MOVZWL #SS$_NORMAL,R0 ;ASSUME NORMAL COMPLETION STATUS BRB FUNCXT ;FUNCTION EXIT FATALERR: ;UNRECOVERABLE ERROR MOVZWL #SS$_DRVERR,R0 ;ASSUME DRIVE ERROR STATUS RESETXFR: ; dummy entry ... should never really get here MOVL UCB$L_IRP(R5),R3 ;GET I/O PKT MNEGW IRP$W_BCNT(R3),UCB$W_BCR(R5) ; RESET BYTECOUNT ; BRW FUNCXT FUNCXT: ;FUNCTION EXIT CLRL R1 ;CLEAR 2ND LONGWORD OF IOSB REQCOM ; COMPLETE REQUEST .PAGE ; PWRFAIL: ;POWER FAILURE BICW #UCB$M_POWER,UCB$W_STS(R5) ;CLEAR POWER FAILURE BIT MOVL UCB$L_IRP(R5),R3 ;GET ADDRESS OF I/O PACKET MOVQ IRP$L_SVAPTE(R3),- ;RESTORE TRANSFER PARAMETERS UCB$L_SVAPTE(R5) ;... BRW vq_STARTIO ;START REQUEST OVER vq_INT:: vq_UNSOLNT:: POPR #^M REI ;DUMMY RETURN FROM ANY INTERRUPT ;; ; FIX SPLITS... ; RETURN IRP TO OUR UCB ADDRESS ; THEN REQCOM ; ; TRICK IS TO GET OUR UCB ADDRESS BACK WHEN WE REGAIN CONTROL. DO SO VIA ; JIGGERY-POKERY WITH THE ADDRESS WE CALL. STORE UCB ADDRESSES IN A TABLE ; INTERNALLY AND USE THE CALL ADDRESS TO GET WHERE WE ARE BACK AGAIN. ; ; Note: On entry, r5 points at the IRP we're to handle. We bash this ; information and regenerate it, since irp$l_ucb has already been ; bashed and can't locate our UCB anyway. Therefore we let the return ; address give us the VE: UCB address implicitly via local save and ; restore; each VE: unit returns to a different entry point which preloads ; r5 with a different offset. Once the ucb is located, ucb$l_irp gets ; back the IRP address. This is possibly extra work; one can imagine ; that some IRP fields like SEGVBN could be used to hold the VE: ucb ; address temporarily, since they are saved/restored internally. This ; would allow some address arithmetic to be dispensed with. The current ; method is merely intended to work and NOT force us to let the host driver ; (and maybe other host software) look at bogus IRP fields, and also ; let us remain blissfully ignorant of how VMS handles IRPEs for purposes ; of this driver. (ecch...having to figure out a way to keep i/o post ; processing out of OUR IRPEs (for sure) while using them for temporary ; storage... what a thought!) ; ; NOTE FOLLOWING CODE ASSUMES vq_UNITS IS 2 OR MORE. V_UNIT=0 V_UNM=1 vq_FXS0:: MOVL I^#V_UNIT,R4 BRW vq_FIXSPLIT ;GO HANDLE vq_FXPL==.-vq_FXS0 ;LENGTH IN BYTES OF THIS LITTLE CODE SEGMENT V_UNIT=V_UNIT+4 ;PASS TO NEXT UNIT .MACRO XVEC LBLC vq_FXS'LBLC: MOVL I^#V_UNIT,R4 BRW vq_FIXSPLIT .ENDM .REPEAT XVEC \V_UNM V_UNIT=V_UNIT+4 ;PASS TO NEXT UNIT V_UNM=V_UNM+1 .ENDR vq_FIXSPLIT: ; GET OLD PID.. ; IN OUR UCB$PPID LONGWORD... ; NOTE!!! PROBABLY NEEDS MODS FOR VMS V5!!! ;some cleanup for host needed here. Note that r5 enters as IRP address. ; Use initial R5 to help reset host's system... movl irp$l_ucb(r5),r3 ;get host's UCB addr movl r5,r2 ;store entry IRP address for check later PUSHL R4 ;NEED TO WORK IN R5 MOVAB vq_UCBTBL,R5 ADDL2 (SP)+,R5 ;R5 NOW POINTS AT UCB ADDRESS MOVL (R5),R5 ;NOW HAVE OUR UCB ADDRESS IN R5 decw ucb$w_qlen(r3) ;cleanup host's q len as ioc$iopost would have bgeq 6$ clrw ucb$w_qlen(r3) ;force queue length zero 6$: cmpl r2,ucb$l_irp(r5) ;got the correct IRP??? beql 7$ ;if eql yes ; MUST avoid screwup where we don't have the correct IRP since there's ; no connection directly between IRP and UCB. VD: unit being busy should ; avoid this error, BUT we have no way to be certain of this w/o exhaustive ; system code checks. .if df,x$$$dt ; oops! wrong IRP. major screwup... jsb g^ini$brk .endc rsb ;else wrong IRP, don't do more damage. 7$: ; notice stack is now clean too. movl r5,r4 FORK ;go fork on our UCB now (VE: ucb) movl r4,r5 MOVL UCB$L_IRP(R5),R3 ; POINT R3 AT IRP AGAIN ; Now ensure we can get to a second container IF we are in multi- ; container mode AND IF this is warranted by current state of ; our I/O... tstl ucb$shmd(r5) ; in dual file shadow mode? bneq 4300$ 4301$: brw 300$ 4300$: ; beql 300$ ; if eql no, go to normal processing ; doing dual shadowing. We must check write... ; (If doing write, we ALWAYS have to write both files (assuming we ; are shadowing at all). If doing read, we just go on and finish ; since we only read from ONE file, never both. tstl ucb$rwdir(r5) ; 0 = read direction; also no problem. beql 4301$ ; if reading, normal processing. ; ok, writing. if ucb$ucbos is #ucb$hucb offset, we just did first ; write of two to be done. If not, we just finished the second of ; two writes and should finish up. cmpl ucb$ucbos(r5),#ucb$hucb ;this the first write? bneq 301$ ; no, just finished second write, so ; finish up. ; Just did FIRST of two writes. Fix things up and set up for the ; SECOND write. movl #ucb$hucb2,ucb$ucbos(r5) ;set up for second pass ; Issue of what to do with errors arises here. Save first host's ; driver status in ucb$er1 field here. movl irp$l_media(r3),ucb$er1(r5) ;store original error MOVL UCB$LMEDIA(R5),IRP$L_MEDIA(R3) ; restore original LBN ; Now dispatch the IRP back to the second driver as physical I/O in ; second driver's host context. ; (We're already at proper fork level to get away with this...) ; Reset byte count also from our saved values movl ucb$q_vq_svaptetmp(r5),irp$l_svapte(r3) ;reset svapte movl ucb$q_vq_svaptetmp+4(r5),irp$l_bcnt(r3) ;and orig bcnt ; movl irp$l_bcnt(r3),irp$l_obcnt(r3) ;and reset to actual movl ucb$owind(r5),irp$l_wind(r3) ;store window ptr movl ucb$osegv(r5),irp$l_segvbn(r3) ;store segment vbn also .if df,x$$$dt ; examine IRP before we queue it to the second host jsb g^ini$brk .endc ADDL2 UCB$HLBN2(R5),IRP$L_MEDIA(R3) ;ADJUST LBN IN IO PKT MOVL UCB$HUCB2(R5),IRP$L_UCB(R3) ;FIX UP PTR IN I/O PKT MOVL UCB$HUCB2(R5),R5 ;NOW POINT AT HOST UCB OURSELVES clrl irp$l_media+4(r3) ;zero old bytes transferred in pass 1 MOVL IRP$L_MEDIA(R3),R0 ;GET LBN TO CONVERT JSB G^IOC$CVTLOGPHY ; LET THE EXEC DO IT JMP G^EXE$INSIOQc ; INSERT PACKET INTO HOST'S QUEUE ; We don't get past here directly. 301$: ; merge error codes of two writes blbs ucb$er1(r5),300$ ; if first write was OK, skip ; error on first of 2 writes. Return an error code somehow. ; Return errors blbs irp$l_media(r3),302$ ;if 2nd write OK rtn ; ss$_ivstsflg movl #ss$_noslot,irp$l_media(r3) ; else set ss$_noslot ; ss$_noslot means both writes bad brb 300$ 302$: movl #ss$_ivstsflg,irp$l_media(r3) ; ss$_ivstsflg means first write failed but second was OK ; ; These errors are unlikely except as a result of errors in writing ; and can be used to flag whether one or both files went wrong. ; (normal error flags return on failures in second write.) 300$: TSTL UCB$PPID(R5) ; ENSURE PID IS NONZERO AS SAVED BEQL 15$ ; SKIP BASH IF NOT MOVL UCB$PPID(R5),IRP$L_PID(R3) ;RESTORE THE OLD PID ; since we may now have later parts of virtual, paging, or swapping I/O ; to do, restore saved byte counts and function codes. ; movl ucb$obct(r5),irp$l_obcnt(r3) ;restore orig byte cnt ; movl ucb$owind(r5),irp$l_wind(r3) ;restore window pointer movl ucb$osegv(r5),irp$l_segvbn(r3) ;restore segment vbn also brb 1501$ 15$: .if df,x$$$dt jsb g^ini$brk .endc clrl irp$l_pid(r3) ;make sure we DON'T get back here anyway! ; this is actually an error condition and should NEVER occur... ; movl ucb$obct(r5),irp$l_obcnt(r3) ;restore orig byte cnt 1501$: CLRL UCB$PPID(R5) ; ZERO SAVED PID FIELD FOR CLEANLINESS MOVL R5,IRP$L_UCB(R3) ;RESTORE VE: AS UCB IN IRP TOO ; GRAB R0 AND R1 AS REQCOM IN HOST DRIVER LEFT THEM... MOVL IRP$L_MEDIA(R3),R0 ;GET BACK R0 MOVL IRP$L_MEDIA+4(R3),R1 ;AND R1 ; R0, R1 ARE AS HOST DRIVER LEFT THEM. R5 POINTS TO CORRECT UCB. ; ===> GO FOR IT !!! ; ; Now restore the original IRP$L_MEDIA field of the IRP in case error ; paths in IOC$REQCOM ever need it. Some very low XQP cache situations ; may occasionally need this, though in reasonable sysgen configs it ; should never be needed. This is the one area that got bashed during ; the earlier I/O completion processing in the host driver. ; MOVL UCB$LMEDIA(R5),IRP$L_MEDIA(R3) ; ; notice that for virtual I/O, the IRP's IRP$L_SEGVBN longword still ; has the starting VIRTUAL block number of the I/O request in the context ; of the virtual disk. This must be present as any second and later parts ; of the I/O request modify that field to compute where to go for the ; next I/O. Due to getting back here, the host driver need never know ; about this; it is basically doing ONLY physical and logical I/O where ; this sort of completion jiggery-pokery does not occur. ; - GCE ; Now go REALLY complete the I/O (possibly causing more I/O and certainly ; ensuring the VE: I/O queue is emptied and VE: unbusied after all is done.) JSB @#IOC$REQCOM ; GO COMPLETE THE I/O REQUEST IN VE: CONTEXT ; (OR DO I/O SPLIT NEXT PART IN VE: CONTEXT!) ; ALSO, RETURN **HERE**, SO WE CAN WRAP UP ALL ELSE. ; .IF NDF,VMS$V5 ; ENBINT ; RESTORE IPL TO IPL$_IOPOST ; .IFF ; FORKUNLOCK NEWIPL=(SP)+ ; .ENDC ; rsb exits the fork level. ; IPL4 level exited at fork above, with stack intact at that point. ; iopost saves/restores regs, so r5 bash is ok. RSB ; GET BACK TO HOST SOMETIME ; BLOCK OF UCB ADDRESSES vq_UCBTBL:: .BLKL vq_UNITS .LONG 0,0 ;SAFETY vq_END: ;ADDRESS OF LAST LOCATION IN DRIVER .END