[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [Xen-changelog] [IA64] update xenivt.S and xenentry.S
# HG changeset patch # User awilliam@xxxxxxxxxxx # Node ID ea181d8577124856ce003b80fedc310532082b34 # Parent 00141f6d15e005969b14776eb87f5e217b5eb851 [IA64] update xenivt.S and xenentry.S Update xenentry.S and xenivt.S for linux 2.6.16.13. Signed-off-by: Tristan Gingold <tristan.gingold@xxxxxxxx> --- linux-2.6-xen-sparse/arch/ia64/xen/xenentry.S | 198 ++++---- linux-2.6-xen-sparse/arch/ia64/xen/xenivt.S | 586 +++++++++++++++----------- 2 files changed, 447 insertions(+), 337 deletions(-) diff -r 00141f6d15e0 -r ea181d857712 linux-2.6-xen-sparse/arch/ia64/xen/xenentry.S --- a/linux-2.6-xen-sparse/arch/ia64/xen/xenentry.S Wed May 10 15:58:36 2006 -0600 +++ b/linux-2.6-xen-sparse/arch/ia64/xen/xenentry.S Wed May 10 17:07:06 2006 -0600 @@ -83,11 +83,7 @@ GLOBAL_ENTRY(ia64_switch_to) mov r8=1 ;; st4 [r27]=r8 // psr.ic back on - ;; -#else -(p6) ssm psr.ic // if we had to map, reenable the psr.ic bit FIRST!!! - ;; -(p6) srlz.d +#else ld8 sp=[r21] // load kernel stack pointer of new task mov IA64_KR(CURRENT)=in0 // update "current" application register #endif @@ -136,6 +132,11 @@ GLOBAL_ENTRY(ia64_switch_to) #endif ;; itr.d dtr[r25]=r23 // wire in new mapping... +#ifndef CONFIG_XEN + ssm psr.ic // reenable the psr.ic bit + ;; + srlz.d +#endif br.cond.sptk .done #ifdef CONFIG_XEN END(xen_switch_to) @@ -216,7 +217,9 @@ GLOBAL_ENTRY(ia64_trace_syscall) .mem.offset 0,0; st8.spill [r2]=r8 // store return value in slot for r8 .mem.offset 8,0; st8.spill [r3]=r10 // clear error indication in slot for r10 br.call.sptk.many rp=syscall_trace_leave // give parent a chance to catch return value -.ret3: br.cond.sptk .work_pending_syscall_end +.ret3: +(pUStk) cmp.eq.unc p6,p0=r0,r0 // p6 <- pUStk + br.cond.sptk .work_pending_syscall_end strace_error: ld8 r3=[r2] // load pt_regs.r8 @@ -246,7 +249,7 @@ END(ia64_trace_syscall) * r8-r11: restored (syscall return value(s)) * r12: restored (user-level stack pointer) * r13: restored (user-level thread pointer) - * r14: cleared + * r14: set to __kernel_syscall_via_epc * r15: restored (syscall #) * r16-r17: cleared * r18: user-level b6 @@ -267,7 +270,7 @@ END(ia64_trace_syscall) * pr: restored (user-level pr) * b0: restored (user-level rp) * b6: restored - * b7: cleared + * b7: set to __kernel_syscall_via_epc * ar.unat: restored (user-level ar.unat) * ar.pfs: restored (user-level ar.pfs) * ar.rsc: restored (user-level ar.rsc) @@ -331,20 +334,20 @@ ENTRY(ia64_leave_syscall) ;; (p6) ld4 r31=[r18] // load current_thread_info()->flags ld8 r19=[r2],PT(B6)-PT(LOADRS) // load ar.rsc value for "loadrs" - mov b7=r0 // clear b7 - ;; - ld8 r23=[r3],PT(R11)-PT(AR_BSPSTORE) // load ar.bspstore (may be garbage) + nop.i 0 + ;; + mov r16=ar.bsp // M2 get existing backing store pointer ld8 r18=[r2],PT(R9)-PT(B6) // load b6 (p6) and r15=TIF_WORK_MASK,r31 // any work other than TIF_SYSCALL_TRACE? ;; - mov r16=ar.bsp // M2 get existing backing store pointer + ld8 r23=[r3],PT(R11)-PT(AR_BSPSTORE) // load ar.bspstore (may be garbage) (p6) cmp4.ne.unc p6,p0=r15, r0 // any special work pending? (p6) br.cond.spnt .work_pending_syscall ;; // start restoring the state saved on the kernel stack (struct pt_regs): ld8 r9=[r2],PT(CR_IPSR)-PT(R9) ld8 r11=[r3],PT(CR_IIP)-PT(R11) - mov f6=f0 // clear f6 +(pNonSys) break 0 // bug check: we shouldn't be here if pNonSys is TRUE! ;; invala // M0|1 invalidate ALAT #ifdef CONFIG_XEN @@ -358,57 +361,68 @@ ENTRY(ia64_leave_syscall) st4 [r29]=r0 // note: clears both vpsr.i and vpsr.ic! ;; #else - rsm psr.i | psr.ic // M2 initiate turning off of interrupt and interruption collection -#endif - mov f9=f0 // clear f9 - - ld8 r29=[r2],16 // load cr.ipsr - ld8 r28=[r3],16 // load cr.iip - mov f8=f0 // clear f8 + rsm psr.i | psr.ic // M2 turn off interrupts and interruption collection +#endif + cmp.eq p9,p0=r0,r0 // A set p9 to indicate that we should restore cr.ifs + + ld8 r29=[r2],16 // M0|1 load cr.ipsr + ld8 r28=[r3],16 // M0|1 load cr.iip + mov r22=r0 // A clear r22 ;; ld8 r30=[r2],16 // M0|1 load cr.ifs - mov.m ar.ssd=r0 // M2 clear ar.ssd - cmp.eq p9,p0=r0,r0 // set p9 to indicate that we should restore cr.ifs - ;; ld8 r25=[r3],16 // M0|1 load ar.unat - mov.m ar.csd=r0 // M2 clear ar.csd - mov r22=r0 // clear r22 +(pUStk) add r14=IA64_TASK_THREAD_ON_USTACK_OFFSET,r13 ;; ld8 r26=[r2],PT(B0)-PT(AR_PFS) // M0|1 load ar.pfs -(pKStk) mov r22=psr // M2 read PSR now that interrupts are disabled - mov f10=f0 // clear f10 - ;; - ld8 r21=[r2],PT(AR_RNAT)-PT(B0) // load b0 - ld8 r27=[r3],PT(PR)-PT(AR_RSC) // load ar.rsc - mov f11=f0 // clear f11 - ;; - ld8 r24=[r2],PT(AR_FPSR)-PT(AR_RNAT) // load ar.rnat (may be garbage) - ld8 r31=[r3],PT(R1)-PT(PR) // load predicates -(pUStk) add r14=IA64_TASK_THREAD_ON_USTACK_OFFSET,r13 - ;; - ld8 r20=[r2],PT(R12)-PT(AR_FPSR) // load ar.fpsr - ld8.fill r1=[r3],16 // load r1 -(pUStk) mov r17=1 - ;; - srlz.d // M0 ensure interruption collection is off - ld8.fill r13=[r3],16 - mov f7=f0 // clear f7 - ;; - ld8.fill r12=[r2] // restore r12 (sp) - ld8.fill r15=[r3] // restore r15 - addl r3=THIS_CPU(ia64_phys_stacked_size_p8),r0 - ;; -(pUStk) ld4 r3=[r3] // r3 = cpu_data->phys_stacked_size_p8 -(pUStk) st1 [r14]=r17 - mov b6=r18 // I0 restore b6 - ;; - mov r14=r0 // clear r14 - shr.u r18=r19,16 // I0|1 get byte size of existing "dirty" partition -(pKStk) br.cond.dpnt.many skip_rbs_switch - - mov.m ar.ccv=r0 // clear ar.ccv -(pNonSys) br.cond.dpnt.many dont_preserve_current_frame - br.cond.sptk.many rbs_switch +(pKStk) mov r22=psr // M2 read PSR now that interrupts are disabled + nop 0 + ;; + ld8 r21=[r2],PT(AR_RNAT)-PT(B0) // M0|1 load b0 + ld8 r27=[r3],PT(PR)-PT(AR_RSC) // M0|1 load ar.rsc + mov f6=f0 // F clear f6 + ;; + ld8 r24=[r2],PT(AR_FPSR)-PT(AR_RNAT) // M0|1 load ar.rnat (may be garbage) + ld8 r31=[r3],PT(R1)-PT(PR) // M0|1 load predicates + mov f7=f0 // F clear f7 + ;; + ld8 r20=[r2],PT(R12)-PT(AR_FPSR) // M0|1 load ar.fpsr + ld8.fill r1=[r3],16 // M0|1 load r1 +(pUStk) mov r17=1 // A + ;; +(pUStk) st1 [r14]=r17 // M2|3 + ld8.fill r13=[r3],16 // M0|1 + mov f8=f0 // F clear f8 + ;; + ld8.fill r12=[r2] // M0|1 restore r12 (sp) + ld8.fill r15=[r3] // M0|1 restore r15 + mov b6=r18 // I0 restore b6 + + addl r17=THIS_CPU(ia64_phys_stacked_size_p8),r0 // A + mov f9=f0 // F clear f9 +(pKStk) br.cond.dpnt.many skip_rbs_switch // B + + srlz.d // M0 ensure interruption collection is off (for cover) + shr.u r18=r19,16 // I0|1 get byte size of existing "dirty" partition +#ifdef CONFIG_XEN + XEN_HYPER_COVER; +#else + cover // B add current frame into dirty partition & set cr.ifs +#endif + ;; +(pUStk) ld4 r17=[r17] // M0|1 r17 = cpu_data->phys_stacked_size_p8 + mov r19=ar.bsp // M2 get new backing store pointer + mov f10=f0 // F clear f10 + + nop.m 0 + movl r14=__kernel_syscall_via_epc // X + ;; + mov.m ar.csd=r0 // M2 clear ar.csd + mov.m ar.ccv=r0 // M2 clear ar.ccv + mov b7=r14 // I0 clear b7 (hint with __kernel_syscall_via_epc) + + mov.m ar.ssd=r0 // M2 clear ar.ssd + mov f11=f0 // F clear f11 + br.cond.sptk.many rbs_switch // B #ifdef CONFIG_XEN END(xen_leave_syscall) #else @@ -546,7 +560,7 @@ GLOBAL_ENTRY(ia64_leave_kernel) ldf.fill f7=[r2],PT(F11)-PT(F7) ldf.fill f8=[r3],32 ;; - srlz.i // ensure interruption collection is off + srlz.d // ensure that inter. collection is off (VHPT is don't care, since text is pinned) mov ar.ccv=r15 ;; ldf.fill f11=[r2] @@ -556,29 +570,29 @@ GLOBAL_ENTRY(ia64_leave_kernel) movl r2=XSI_BANK1_R16 movl r3=XSI_BANK1_R16+8 ;; - st8.spill [r2]=r16,16 - st8.spill [r3]=r17,16 - ;; - st8.spill [r2]=r18,16 - st8.spill [r3]=r19,16 - ;; - st8.spill [r2]=r20,16 - st8.spill [r3]=r21,16 - ;; - st8.spill [r2]=r22,16 - st8.spill [r3]=r23,16 - ;; - st8.spill [r2]=r24,16 - st8.spill [r3]=r25,16 - ;; - st8.spill [r2]=r26,16 - st8.spill [r3]=r27,16 - ;; - st8.spill [r2]=r28,16 - st8.spill [r3]=r29,16 - ;; - st8.spill [r2]=r30,16 - st8.spill [r3]=r31,16 +.mem.offset 0,0; st8.spill [r2]=r16,16 +.mem.offset 8,0; st8.spill [r3]=r17,16 + ;; +.mem.offset 0,0; st8.spill [r2]=r18,16 +.mem.offset 8,0; st8.spill [r3]=r19,16 + ;; +.mem.offset 0,0; st8.spill [r2]=r20,16 +.mem.offset 8,0; st8.spill [r3]=r21,16 + ;; +.mem.offset 0,0; st8.spill [r2]=r22,16 +.mem.offset 8,0; st8.spill [r3]=r23,16 + ;; +.mem.offset 0,0; st8.spill [r2]=r24,16 +.mem.offset 8,0; st8.spill [r3]=r25,16 + ;; +.mem.offset 0,0; st8.spill [r2]=r26,16 +.mem.offset 8,0; st8.spill [r3]=r27,16 + ;; +.mem.offset 0,0; st8.spill [r2]=r28,16 +.mem.offset 8,0; st8.spill [r3]=r29,16 + ;; +.mem.offset 0,0; st8.spill [r2]=r30,16 +.mem.offset 8,0; st8.spill [r3]=r31,16 ;; movl r2=XSI_BANKNUM;; st4 [r2]=r0; @@ -641,14 +655,14 @@ GLOBAL_ENTRY(ia64_leave_kernel) */ (pNonSys) br.cond.dpnt dont_preserve_current_frame +#ifdef CONFIG_XEN + XEN_HYPER_COVER; +#else + cover // add current frame into dirty partition and set cr.ifs +#endif + ;; + mov r19=ar.bsp // get new backing store pointer rbs_switch: -#ifdef CONFIG_XEN - XEN_HYPER_COVER; -#else - cover // add current frame into dirty partition and set cr.ifs -#endif - ;; - mov r19=ar.bsp // get new backing store pointer sub r16=r16,r18 // krbs = old bsp - size of dirty partition cmp.ne p9,p0=r0,r0 // clear p9 to skip restore of cr.ifs ;; @@ -723,14 +737,14 @@ rse_clear_invalid: mov loc5=0 mov loc6=0 mov loc7=0 -(pRecurse) br.call.sptk.few b0=rse_clear_invalid +(pRecurse) br.call.dptk.few b0=rse_clear_invalid ;; mov loc8=0 mov loc9=0 cmp.ne pReturn,p0=r0,in1 // if recursion count != 0, we need to do a br.ret mov loc10=0 mov loc11=0 -(pReturn) br.ret.sptk.many b0 +(pReturn) br.ret.dptk.many b0 #endif /* !CONFIG_ITANIUM */ # undef pRecurse # undef pReturn diff -r 00141f6d15e0 -r ea181d857712 linux-2.6-xen-sparse/arch/ia64/xen/xenivt.S --- a/linux-2.6-xen-sparse/arch/ia64/xen/xenivt.S Wed May 10 15:58:36 2006 -0600 +++ b/linux-2.6-xen-sparse/arch/ia64/xen/xenivt.S Wed May 10 17:07:06 2006 -0600 @@ -87,16 +87,17 @@ ENTRY(vhpt_miss) * (the "original") TLB miss, which may either be caused by an instruction * fetch or a data access (or non-access). * - * What we do here is normal TLB miss handing for the _original_ miss, followed - * by inserting the TLB entry for the virtual page table page that the VHPT - * walker was attempting to access. The latter gets inserted as long - * as both L1 and L2 have valid mappings for the faulting address. - * The TLB entry for the original miss gets inserted only if - * the L3 entry indicates that the page is present. + * What we do here is normal TLB miss handing for the _original_ miss, + * followed by inserting the TLB entry for the virtual page table page + * that the VHPT walker was attempting to access. The latter gets + * inserted as long as page table entry above pte level have valid + * mappings for the faulting address. The TLB entry for the original + * miss gets inserted only if the pte entry indicates that the page is + * present. * * do_page_fault gets invoked in the following cases: * - the faulting virtual address uses unimplemented address bits - * - the faulting virtual address has no L1, L2, or L3 mapping + * - the faulting virtual address has no valid page table mapping */ #ifdef CONFIG_XEN movl r16=XSI_IFA @@ -127,7 +128,7 @@ ENTRY(vhpt_miss) shl r21=r16,3 // shift bit 60 into sign bit shr.u r17=r16,61 // get the region number into r17 ;; - shr r22=r21,3 + shr.u r22=r21,3 #ifdef CONFIG_HUGETLB_PAGE extr.u r26=r25,2,6 ;; @@ -139,7 +140,7 @@ ENTRY(vhpt_miss) #endif ;; cmp.eq p6,p7=5,r17 // is IFA pointing into to region 5? - shr.u r18=r22,PGDIR_SHIFT // get bits 33-63 of the faulting address + shr.u r18=r22,PGDIR_SHIFT // get bottom portion of pgd index bit ;; (p7) dep r17=r17,r19,(PAGE_SHIFT-3),3 // put region number bits in place @@ -150,41 +151,54 @@ ENTRY(vhpt_miss) (p6) shr.u r21=r21,PGDIR_SHIFT+PAGE_SHIFT (p7) shr.u r21=r21,PGDIR_SHIFT+PAGE_SHIFT-3 ;; -(p6) dep r17=r18,r19,3,(PAGE_SHIFT-3) // r17=PTA + IFA(33,42)*8 -(p7) dep r17=r18,r17,3,(PAGE_SHIFT-6) // r17=PTA + (((IFA(61,63) << 7) | IFA(33,39))*8) +(p6) dep r17=r18,r19,3,(PAGE_SHIFT-3) // r17=pgd_offset for region 5 +(p7) dep r17=r18,r17,3,(PAGE_SHIFT-6) // r17=pgd_offset for region[0-4] cmp.eq p7,p6=0,r21 // unused address bits all zeroes? - shr.u r18=r22,PMD_SHIFT // shift L2 index into position - ;; - ld8 r17=[r17] // fetch the L1 entry (may be 0) - ;; -(p7) cmp.eq p6,p7=r17,r0 // was L1 entry NULL? - dep r17=r18,r17,3,(PAGE_SHIFT-3) // compute address of L2 page table entry - ;; -(p7) ld8 r20=[r17] // fetch the L2 entry (may be 0) - shr.u r19=r22,PAGE_SHIFT // shift L3 index into position - ;; -(p7) cmp.eq.or.andcm p6,p7=r20,r0 // was L2 entry NULL? - dep r21=r19,r20,3,(PAGE_SHIFT-3) // compute address of L3 page table entry - ;; -#ifdef CONFIG_XEN -(p7) ld8 r18=[r21] // read the L3 PTE +#ifdef CONFIG_PGTABLE_4 + shr.u r28=r22,PUD_SHIFT // shift pud index into position +#else + shr.u r18=r22,PMD_SHIFT // shift pmd index into position +#endif + ;; + ld8 r17=[r17] // get *pgd (may be 0) + ;; +(p7) cmp.eq p6,p7=r17,r0 // was pgd_present(*pgd) == NULL? +#ifdef CONFIG_PGTABLE_4 + dep r28=r28,r17,3,(PAGE_SHIFT-3) // r28=pud_offset(pgd,addr) + ;; + shr.u r18=r22,PMD_SHIFT // shift pmd index into position +(p7) ld8 r29=[r28] // get *pud (may be 0) + ;; +(p7) cmp.eq.or.andcm p6,p7=r29,r0 // was pud_present(*pud) == NULL? + dep r17=r18,r29,3,(PAGE_SHIFT-3) // r17=pmd_offset(pud,addr) +#else + dep r17=r18,r17,3,(PAGE_SHIFT-3) // r17=pmd_offset(pgd,addr) +#endif + ;; +(p7) ld8 r20=[r17] // get *pmd (may be 0) + shr.u r19=r22,PAGE_SHIFT // shift pte index into position + ;; +(p7) cmp.eq.or.andcm p6,p7=r20,r0 // was pmd_present(*pmd) == NULL? + dep r21=r19,r20,3,(PAGE_SHIFT-3) // r21=pte_offset(pmd,addr) + ;; +(p7) ld8 r18=[r21] // read *pte +#ifdef CONFIG_XEN movl r19=XSI_ISR ;; ld8 r19=[r19] +#else + mov r19=cr.isr // cr.isr bit 32 tells us if this is an insn miss +#endif ;; (p7) tbit.z p6,p7=r18,_PAGE_P_BIT // page present bit cleared? +#ifdef CONFIG_XEN movl r22=XSI_IHA ;; ld8 r22=[r22] - ;; -#else -(p7) ld8 r18=[r21] // read the L3 PTE - mov r19=cr.isr // cr.isr bit 0 tells us if this is an insn miss - ;; -(p7) tbit.z p6,p7=r18,_PAGE_P_BIT // page present bit cleared? +#else mov r22=cr.iha // get the VHPT address that caused the TLB miss +#endif ;; // avoid RAW on p7 -#endif (p7) tbit.nz.unc p10,p11=r19,32 // is it an instruction TLB miss? dep r23=0,r20,0,PAGE_SHIFT // clear low bits to get page address ;; @@ -198,16 +212,17 @@ ENTRY(vhpt_miss) ;; mov r8=r24 ;; -(p6) br.cond.spnt.many page_fault // handle bad address/page not present (page fault) - ;; - movl r24=XSI_IFA - ;; - st8 [r24]=r22 - ;; #else (p10) itc.i r18 // insert the instruction TLB entry (p11) itc.d r18 // insert the data TLB entry +#endif (p6) br.cond.spnt.many page_fault // handle bad address/page not present (page fault) +#ifdef CONFIG_XEN + movl r24=XSI_IFA + ;; + st8 [r24]=r22 + ;; +#else mov cr.ifa=r22 #endif @@ -242,25 +257,41 @@ ENTRY(vhpt_miss) dv_serialize_data /* - * Re-check L2 and L3 pagetable. If they changed, we may have received a ptc.g + * Re-check pagetable entry. If they changed, we may have received a ptc.g * between reading the pagetable and the "itc". If so, flush the entry we - * inserted and retry. - */ - ld8 r25=[r21] // read L3 PTE again - ld8 r26=[r17] // read L2 entry again - ;; - cmp.ne p6,p7=r26,r20 // did L2 entry change + * inserted and retry. At this point, we have: + * + * r28 = equivalent of pud_offset(pgd, ifa) + * r17 = equivalent of pmd_offset(pud, ifa) + * r21 = equivalent of pte_offset(pmd, ifa) + * + * r29 = *pud + * r20 = *pmd + * r18 = *pte + */ + ld8 r25=[r21] // read *pte again + ld8 r26=[r17] // read *pmd again +#ifdef CONFIG_PGTABLE_4 + ld8 r19=[r28] // read *pud again +#endif + cmp.ne p6,p7=r0,r0 + ;; + cmp.ne.or.andcm p6,p7=r26,r20 // did *pmd change +#ifdef CONFIG_PGTABLE_4 + cmp.ne.or.andcm p6,p7=r19,r29 // did *pud change +#endif mov r27=PAGE_SHIFT<<2 ;; (p6) ptc.l r22,r27 // purge PTE page translation -(p7) cmp.ne.or.andcm p6,p7=r25,r18 // did L3 PTE change +(p7) cmp.ne.or.andcm p6,p7=r25,r18 // did *pte change ;; (p6) ptc.l r16,r27 // purge translation #endif mov pr=r31,-1 // restore predicate registers #ifdef CONFIG_XEN - XEN_HYPER_RFI; + XEN_HYPER_RFI + dv_serialize_data #else rfi #endif @@ -272,10 +303,10 @@ ENTRY(itlb_miss) ENTRY(itlb_miss) DBG_FAULT(1) /* - * The ITLB handler accesses the L3 PTE via the virtually mapped linear + * The ITLB handler accesses the PTE via the virtually mapped linear * page table. If a nested TLB miss occurs, we switch into physical - * mode, walk the page table, and then re-execute the L3 PTE read - * and go on normally after that. + * mode, walk the page table, and then re-execute the PTE read and + * go on normally after that. */ #ifdef CONFIG_XEN movl r16=XSI_IFA @@ -292,11 +323,11 @@ ENTRY(itlb_miss) ;; ld8 r17=[r17] // get virtual address of L3 PTE #else - mov r17=cr.iha // get virtual address of L3 PTE + mov r17=cr.iha // get virtual address of PTE #endif movl r30=1f // load nested fault continuation point ;; -1: ld8 r18=[r17] // read L3 PTE +1: ld8 r18=[r17] // read *pte ;; mov b0=r29 tbit.z p6,p0=r18,_PAGE_P_BIT // page present bit cleared? @@ -320,7 +351,7 @@ 1: ld8 r18=[r17] // read L3 PTE */ dv_serialize_data - ld8 r19=[r17] // read L3 PTE again and see if same + ld8 r19=[r17] // read *pte again and see if same mov r20=PAGE_SHIFT<<2 // setup page size for purge ;; cmp.ne p7,p0=r18,r19 @@ -329,7 +360,8 @@ 1: ld8 r18=[r17] // read L3 PTE #endif mov pr=r31,-1 #ifdef CONFIG_XEN - XEN_HYPER_RFI; + XEN_HYPER_RFI + dv_serialize_data #else rfi #endif @@ -341,10 +373,10 @@ ENTRY(dtlb_miss) ENTRY(dtlb_miss) DBG_FAULT(2) /* - * The DTLB handler accesses the L3 PTE via the virtually mapped linear + * The DTLB handler accesses the PTE via the virtually mapped linear * page table. If a nested TLB miss occurs, we switch into physical - * mode, walk the page table, and then re-execute the L3 PTE read - * and go on normally after that. + * mode, walk the page table, and then re-execute the PTE read and + * go on normally after that. */ #ifdef CONFIG_XEN movl r16=XSI_IFA @@ -361,11 +393,11 @@ dtlb_fault: ;; ld8 r17=[r17] // get virtual address of L3 PTE #else - mov r17=cr.iha // get virtual address of L3 PTE + mov r17=cr.iha // get virtual address of PTE #endif movl r30=1f // load nested fault continuation point ;; -1: ld8 r18=[r17] // read L3 PTE +1: ld8 r18=[r17] // read *pte ;; mov b0=r29 tbit.z p6,p0=r18,_PAGE_P_BIT // page present bit cleared? @@ -390,7 +422,7 @@ 1: ld8 r18=[r17] // read L3 PTE */ dv_serialize_data - ld8 r19=[r17] // read L3 PTE again and see if same + ld8 r19=[r17] // read *pte again and see if same mov r20=PAGE_SHIFT<<2 // setup page size for purge ;; cmp.ne p7,p0=r18,r19 @@ -399,7 +431,8 @@ 1: ld8 r18=[r17] // read L3 PTE #endif mov pr=r31,-1 #ifdef CONFIG_XEN - XEN_HYPER_RFI; + XEN_HYPER_RFI + dv_serialize_data #else rfi #endif @@ -416,19 +449,15 @@ ENTRY(alt_itlb_miss) ld8 r21=[r31],XSI_IFA-XSI_IPSR // get ipsr, point to ifa movl r17=PAGE_KERNEL ;; - movl r19=(((1 << IA64_MAX_PHYS_BITS) - 1) & ~0xfff) - ;; ld8 r16=[r31] // get ifa - mov r31=pr - ;; #else mov r16=cr.ifa // get address that caused the TLB miss movl r17=PAGE_KERNEL mov r21=cr.ipsr +#endif movl r19=(((1 << IA64_MAX_PHYS_BITS) - 1) & ~0xfff) mov r31=pr ;; -#endif #ifdef CONFIG_DISABLE_VHPT shr.u r22=r16,61 // get the region number into r21 ;; @@ -486,17 +515,15 @@ ENTRY(alt_dtlb_miss) movl r19=(((1 << IA64_MAX_PHYS_BITS) - 1) & ~0xfff) ;; ld8 r16=[r31] // get ifa - mov r31=pr - ;; #else mov r16=cr.ifa // get address that caused the TLB miss movl r17=PAGE_KERNEL mov r20=cr.isr movl r19=(((1 << IA64_MAX_PHYS_BITS) - 1) & ~0xfff) mov r21=cr.ipsr +#endif mov r31=pr ;; -#endif #ifdef CONFIG_DISABLE_VHPT shr.u r22=r16,61 // get the region number into r21 ;; @@ -565,12 +592,12 @@ ENTRY(nested_dtlb_miss) * r30: continuation address * r31: saved pr * - * Output: r17: physical address of L3 PTE of faulting address + * Output: r17: physical address of PTE of faulting address * r29: saved b0 * r30: continuation address * r31: saved pr * - * Clobbered: b0, r18, r19, r21, psr.dt (cleared) + * Clobbered: b0, r18, r19, r21, r22, psr.dt (cleared) */ #ifdef CONFIG_XEN XEN_HYPER_RSM_PSR_DT; @@ -579,12 +606,23 @@ ENTRY(nested_dtlb_miss) #endif mov r19=IA64_KR(PT_BASE) // get the page table base address shl r21=r16,3 // shift bit 60 into sign bit +#ifdef CONFIG_XEN + movl r18=XSI_ITIR + ;; + ld8 r18=[r18] +#else + mov r18=cr.itir +#endif ;; shr.u r17=r16,61 // get the region number into r17 + extr.u r18=r18,2,6 // get the faulting page size ;; cmp.eq p6,p7=5,r17 // is faulting address in region 5? - shr.u r18=r16,PGDIR_SHIFT // get bits 33-63 of faulting address - ;; + add r22=-PAGE_SHIFT,r18 // adjustment for hugetlb address + add r18=PGDIR_SHIFT-PAGE_SHIFT,r18 + ;; + shr.u r22=r16,r22 + shr.u r18=r16,r18 (p7) dep r17=r17,r19,(PAGE_SHIFT-3),3 // put region number bits in place srlz.d @@ -594,21 +632,33 @@ ENTRY(nested_dtlb_miss) (p6) shr.u r21=r21,PGDIR_SHIFT+PAGE_SHIFT (p7) shr.u r21=r21,PGDIR_SHIFT+PAGE_SHIFT-3 ;; -(p6) dep r17=r18,r19,3,(PAGE_SHIFT-3) // r17=PTA + IFA(33,42)*8 -(p7) dep r17=r18,r17,3,(PAGE_SHIFT-6) // r17=PTA + (((IFA(61,63) << 7) | IFA(33,39))*8) +(p6) dep r17=r18,r19,3,(PAGE_SHIFT-3) // r17=pgd_offset for region 5 +(p7) dep r17=r18,r17,3,(PAGE_SHIFT-6) // r17=pgd_offset for region[0-4] cmp.eq p7,p6=0,r21 // unused address bits all zeroes? - shr.u r18=r16,PMD_SHIFT // shift L2 index into position - ;; - ld8 r17=[r17] // fetch the L1 entry (may be 0) - ;; -(p7) cmp.eq p6,p7=r17,r0 // was L1 entry NULL? - dep r17=r18,r17,3,(PAGE_SHIFT-3) // compute address of L2 page table entry - ;; -(p7) ld8 r17=[r17] // fetch the L2 entry (may be 0) - shr.u r19=r16,PAGE_SHIFT // shift L3 index into position - ;; -(p7) cmp.eq.or.andcm p6,p7=r17,r0 // was L2 entry NULL? - dep r17=r19,r17,3,(PAGE_SHIFT-3) // compute address of L3 page table entry +#ifdef CONFIG_PGTABLE_4 + shr.u r18=r22,PUD_SHIFT // shift pud index into position +#else + shr.u r18=r22,PMD_SHIFT // shift pmd index into position +#endif + ;; + ld8 r17=[r17] // get *pgd (may be 0) + ;; +(p7) cmp.eq p6,p7=r17,r0 // was pgd_present(*pgd) == NULL? + dep r17=r18,r17,3,(PAGE_SHIFT-3) // r17=p[u|m]d_offset(pgd,addr) + ;; +#ifdef CONFIG_PGTABLE_4 +(p7) ld8 r17=[r17] // get *pud (may be 0) + shr.u r18=r22,PMD_SHIFT // shift pmd index into position + ;; +(p7) cmp.eq.or.andcm p6,p7=r17,r0 // was pud_present(*pud) == NULL? + dep r17=r18,r17,3,(PAGE_SHIFT-3) // r17=pmd_offset(pud,addr) + ;; +#endif +(p7) ld8 r17=[r17] // get *pmd (may be 0) + shr.u r19=r22,PAGE_SHIFT // shift pte index into position + ;; +(p7) cmp.eq.or.andcm p6,p7=r17,r0 // was pmd_present(*pmd) == NULL? + dep r17=r19,r17,3,(PAGE_SHIFT-3) // r17=pte_offset(pmd,addr); (p6) br.cond.spnt page_fault mov b0=r30 br.sptk.many b0 // return to continuation point @@ -626,7 +676,7 @@ END(ikey_miss) // call do_page_fault (predicates are in r31, psr.dt may be off, r16 is faulting address) ENTRY(page_fault) #ifdef CONFIG_XEN - XEN_HYPER_SSM_PSR_DT; + XEN_HYPER_SSM_PSR_DT #else ssm psr.dt ;; @@ -742,11 +792,12 @@ 1: ld8 r18=[r17] ;; // avoid RAW on r18 mov ar.ccv=r18 // set compare value for cmpxchg or r25=_PAGE_D|_PAGE_A,r18 // set the dirty and accessed bits - ;; - cmpxchg8.acq r26=[r17],r25,ar.ccv + tbit.z p7,p6 = r18,_PAGE_P_BIT // Check present bit + ;; +(p6) cmpxchg8.acq r26=[r17],r25,ar.ccv // Only update if page is present mov r24=PAGE_SHIFT<<2 ;; - cmp.eq p6,p7=r26,r18 +(p6) cmp.eq p6,p7=r26,r18 // Only compare if page is present ;; (p6) itc.d r25 // install updated PTE ;; @@ -775,7 +826,8 @@ 1: ld8 r18=[r17] #endif mov pr=r31,-1 // restore pr #ifdef CONFIG_XEN - XEN_HYPER_RFI; + XEN_HYPER_RFI + dv_serialize_data #else rfi #endif @@ -826,11 +878,12 @@ 1: ld8 r18=[r17] ;; mov ar.ccv=r18 // set compare value for cmpxchg or r25=_PAGE_A,r18 // set the accessed bit - ;; - cmpxchg8.acq r26=[r17],r25,ar.ccv + tbit.z p7,p6 = r18,_PAGE_P_BIT // Check present bit + ;; +(p6) cmpxchg8.acq r26=[r17],r25,ar.ccv // Only if page present mov r24=PAGE_SHIFT<<2 ;; - cmp.eq p6,p7=r26,r18 +(p6) cmp.eq p6,p7=r26,r18 // Only if page present ;; #ifdef CONFIG_XEN mov r26=r8 @@ -869,7 +922,8 @@ 1: ld8 r18=[r17] #endif /* !CONFIG_SMP */ mov pr=r31,-1 #ifdef CONFIG_XEN - XEN_HYPER_RFI; + XEN_HYPER_RFI + dv_serialize_data #else rfi #endif @@ -892,11 +946,13 @@ ENTRY(daccess_bit) movl r30=1f // load continuation point in case of nested fault ;; #ifdef CONFIG_XEN - mov r18=r8; - mov r8=r16; - XEN_HYPER_THASH;; - mov r17=r8; - mov r8=r18;; + mov r18=r8 + mov r8=r16 + XEN_HYPER_THASH + ;; + mov r17=r8 + mov r8=r18 + ;; #else thash r17=r16 // compute virtual address of L3 PTE #endif @@ -909,11 +965,12 @@ 1: ld8 r18=[r17] ;; // avoid RAW on r18 mov ar.ccv=r18 // set compare value for cmpxchg or r25=_PAGE_A,r18 // set the dirty bit - ;; - cmpxchg8.acq r26=[r17],r25,ar.ccv + tbit.z p7,p6 = r18,_PAGE_P_BIT // Check present bit + ;; +(p6) cmpxchg8.acq r26=[r17],r25,ar.ccv // Only if page is present mov r24=PAGE_SHIFT<<2 ;; - cmp.eq p6,p7=r26,r18 +(p6) cmp.eq p6,p7=r26,r18 // Only if page is present ;; #ifdef CONFIG_XEN mov r26=r8 @@ -950,7 +1007,8 @@ 1: ld8 r18=[r17] mov b0=r29 // restore b0 mov pr=r31,-1 #ifdef CONFIG_XEN - XEN_HYPER_RFI; + XEN_HYPER_RFI + dv_serialize_data #else rfi #endif @@ -976,143 +1034,157 @@ ENTRY(break_fault) * to prevent leaking bits from kernel to user level. */ DBG_FAULT(11) - mov r16=IA64_KR(CURRENT) // r16 = current task; 12 cycle read lat. -#ifdef CONFIG_XEN - movl r31=XSI_IPSR - ;; - ld8 r29=[r31],XSI_IIP-XSI_IPSR // get ipsr, point to iip - mov r18=__IA64_BREAK_SYSCALL - mov r21=ar.fpsr - ;; - ld8 r28=[r31],XSI_IIM-XSI_IIP // get iip, point to iim - mov r19=b6 - mov r25=ar.unat - ;; - ld8 r17=[r31] // get iim - mov r27=ar.rsc - mov r26=ar.pfs - ;; -#else - mov r17=cr.iim - mov r18=__IA64_BREAK_SYSCALL - mov r21=ar.fpsr - mov r29=cr.ipsr - mov r19=b6 - mov r25=ar.unat - mov r27=ar.rsc - mov r26=ar.pfs - mov r28=cr.iip -#endif - mov r31=pr // prepare to save predicates - mov r20=r1 - ;; + mov.m r16=IA64_KR(CURRENT) // M2 r16 <- current task (12 cyc) +#ifdef CONFIG_XEN + movl r22=XSI_IPSR + ;; + ld8 r29=[r22],XSI_IIM-XSI_IPSR // get ipsr, point to iip +#else + mov r29=cr.ipsr // M2 (12 cyc) +#endif + mov r31=pr // I0 (2 cyc) + +#ifdef CONFIG_XEN + ;; + ld8 r17=[r22],XSI_IIP-XSI_IIM +#else + mov r17=cr.iim // M2 (2 cyc) +#endif + mov.m r27=ar.rsc // M2 (12 cyc) + mov r18=__IA64_BREAK_SYSCALL // A + + mov.m ar.rsc=0 // M2 + mov.m r21=ar.fpsr // M2 (12 cyc) + mov r19=b6 // I0 (2 cyc) + ;; + mov.m r23=ar.bspstore // M2 (12 cyc) + mov.m r24=ar.rnat // M2 (5 cyc) + mov.i r26=ar.pfs // I0 (2 cyc) + + invala // M0|1 + nop.m 0 // M + mov r20=r1 // A save r1 + + nop.m 0 + movl r30=sys_call_table // X + +#ifdef CONFIG_XEN + ld8 r28=[r22] +#else + mov r28=cr.iip // M2 (2 cyc) +#endif + cmp.eq p0,p7=r18,r17 // I0 is this a system call? +(p7) br.cond.spnt non_syscall // B no -> + // + // From this point on, we are definitely on the syscall-path + // and we can use (non-banked) scratch registers. + // +/////////////////////////////////////////////////////////////////////// + mov r1=r16 // A move task-pointer to "addl"-addressable reg + mov r2=r16 // A setup r2 for ia64_syscall_setup + add r9=TI_FLAGS+IA64_TASK_SIZE,r16 // A r9 = ¤t_thread_info()->flags + adds r16=IA64_TASK_THREAD_ON_USTACK_OFFSET,r16 - cmp.eq p0,p7=r18,r17 // is this a system call? (p7 <- false, if so) -(p7) br.cond.spnt non_syscall - ;; - ld1 r17=[r16] // load current->thread.on_ustack flag - st1 [r16]=r0 // clear current->thread.on_ustack flag - add r1=-IA64_TASK_THREAD_ON_USTACK_OFFSET,r16 // set r1 for MINSTATE_START_SAVE_MIN_VIRT - ;; - invala - - /* adjust return address so we skip over the break instruction: */ - - extr.u r8=r29,41,2 // extract ei field from cr.ipsr - ;; - cmp.eq p6,p7=2,r8 // isr.ei==2? - mov r2=r1 // setup r2 for ia64_syscall_setup - ;; -(p6) mov r8=0 // clear ei to 0 -(p6) adds r28=16,r28 // switch cr.iip to next bundle cr.ipsr.ei wrapped -(p7) adds r8=1,r8 // increment ei to next slot - ;; - cmp.eq pKStk,pUStk=r0,r17 // are we in kernel mode already? - dep r29=r8,r29,41,2 // insert new ei into cr.ipsr - ;; - - // switch from user to kernel RBS: - MINSTATE_START_SAVE_MIN_VIRT - br.call.sptk.many b7=ia64_syscall_setup - ;; + adds r15=-1024,r15 // A subtract 1024 from syscall number + mov r3=NR_syscalls - 1 + ;; + ld1.bias r17=[r16] // M0|1 r17 = current->thread.on_ustack flag + ld4 r9=[r9] // M0|1 r9 = current_thread_info()->flags + extr.u r8=r29,41,2 // I0 extract ei field from cr.ipsr + + shladd r30=r15,3,r30 // A r30 = sys_call_table + 8*(syscall-1024) + addl r22=IA64_RBS_OFFSET,r1 // A compute base of RBS + cmp.leu p6,p7=r15,r3 // A syscall number in range? + ;; + + lfetch.fault.excl.nt1 [r22] // M0|1 prefetch RBS +(p6) ld8 r30=[r30] // M0|1 load address of syscall entry point + tnat.nz.or p7,p0=r15 // I0 is syscall nr a NaT? + + mov.m ar.bspstore=r22 // M2 switch to kernel RBS + cmp.eq p8,p9=2,r8 // A isr.ei==2? + ;; + +(p8) mov r8=0 // A clear ei to 0 +(p7) movl r30=sys_ni_syscall // X + +(p8) adds r28=16,r28 // A switch cr.iip to next bundle +(p9) adds r8=1,r8 // A increment ei to next slot + nop.i 0 + ;; + + mov.m r25=ar.unat // M2 (5 cyc) + dep r29=r8,r29,41,2 // I0 insert new ei into cr.ipsr + adds r15=1024,r15 // A restore original syscall number + // + // If any of the above loads miss in L1D, we'll stall here until + // the data arrives. + // +/////////////////////////////////////////////////////////////////////// + st1 [r16]=r0 // M2|3 clear current->thread.on_ustack flag + mov b6=r30 // I0 setup syscall handler branch reg early + cmp.eq pKStk,pUStk=r0,r17 // A were we on kernel stacks already? + + and r9=_TIF_SYSCALL_TRACEAUDIT,r9 // A mask trace or audit + mov r18=ar.bsp // M2 (12 cyc) +(pKStk) br.cond.spnt .break_fixup // B we're already in kernel-mode -- fix up RBS + ;; +.back_from_break_fixup: +(pUStk) addl r1=IA64_STK_OFFSET-IA64_PT_REGS_SIZE,r1 // A compute base of memory stack + cmp.eq p14,p0=r9,r0 // A are syscalls being traced/audited? + br.call.sptk.many b7=ia64_syscall_setup // B +1: + mov ar.rsc=0x3 // M2 set eager mode, pl 0, LE, loadrs=0 + nop 0 #ifdef CONFIG_XEN mov r2=b0; br.call.sptk b0=xen_bsw1;; mov b0=r2;; #else - MINSTATE_END_SAVE_MIN_VIRT // switch to bank 1 -#endif -#ifdef CONFIG_XEN - movl r3=XSI_PSR_IC - mov r16=1 - ;; -#if 1 - st4 [r3]=r16,XSI_PSR_I_ADDR-XSI_PSR_IC // vpsr.ic = 1 - ;; -(p15) ld8 r3=[r3] - ;; -(p15) st1 [r3]=r0,XSI_PEND-XSI_PSR_I_ADDR // if (p15) vpsr.i = 1 - mov r16=r0 - ;; -(p15) ld4 r16=[r3] // if (pending_interrupts) - ;; - cmp.ne p6,p0=r16,r0 + bsw.1 // B (6 cyc) regs are saved, switch to bank 1 +#endif + ;; + +#ifdef CONFIG_XEN + movl r16=XSI_PSR_IC + mov r3=1 + ;; + st4 [r16]=r3,XSI_PSR_I_ADDR-XSI_PSR_IC // vpsr.ic = 1 +#else + ssm psr.ic | PSR_DEFAULT_BITS // M2 now it's safe to re-enable intr.-collection +#endif + movl r3=ia64_ret_from_syscall // X + ;; + + srlz.i // M0 ensure interruption collection is on + mov rp=r3 // I0 set the real return addr +(p10) br.cond.spnt.many ia64_ret_from_syscall // B return if bad call-frame or r15 is a NaT + +#ifdef CONFIG_XEN +(p15) ld8 r16=[r16] // vpsr.i + ;; +(p15) st1 [r16]=r0,XSI_PEND-XSI_PSR_I_ADDR // if (p15) vpsr.i = 1 + mov r2=r0 + ;; +(p15) ld4 r2=[r16] // if (pending_interrupts) + ;; + cmp.ne p6,p0=r2,r0 ;; (p6) ssm psr.i // do a real ssm psr.i - ;; -#else -// st4 [r3]=r16,XSI_PSR_I_ADDR-XSI_PSR_IC // vpsr.ic = 1 - adds r3=XSI_PSR_I_ADDR-XSI_PSR_IC,r3 // SKIP vpsr.ic = 1 - ;; -(p15) ld8 r3=[r3] - ;; -(p15) st1 [r3]=r0,XSI_PEND-XSI_PSR_I_ADDR // if (p15) vpsr.i = 1 - mov r16=r0 - ;; -(p15) ld4 r16=[r3] // if (pending_interrupts) - ;; - cmp.ne p6,p0=r16,r0 - ;; -//(p6) ssm psr.i // do a real ssm psr.i -//(p6) XEN_HYPER_SSM_I; -(p6) break 0x7; - ;; -#endif - mov r3=NR_syscalls - 1 - ;; -#else - ssm psr.ic | PSR_DEFAULT_BITS - ;; - srlz.i // guarantee that interruption collection is on - mov r3=NR_syscalls - 1 - ;; -(p15) ssm psr.i // restore psr.i -#endif - // p10==true means out registers are more than 8 or r15's Nat is true -(p10) br.cond.spnt.many ia64_ret_from_syscall - ;; - movl r16=sys_call_table - - adds r15=-1024,r15 // r15 contains the syscall number---subtract 1024 - movl r2=ia64_ret_from_syscall - ;; - shladd r20=r15,3,r16 // r20 = sys_call_table + 8*(syscall-1024) - cmp.leu p6,p7=r15,r3 // (syscall > 0 && syscall < 1024 + NR_syscalls) ? - mov rp=r2 // set the real return addr - ;; -(p6) ld8 r20=[r20] // load address of syscall entry point -(p7) movl r20=sys_ni_syscall - - add r2=TI_FLAGS+IA64_TASK_SIZE,r13 - ;; - ld4 r2=[r2] // r2 = current_thread_info()->flags - ;; - and r2=_TIF_SYSCALL_TRACEAUDIT,r2 // mask trace or audit - ;; - cmp.eq p8,p0=r2,r0 - mov b6=r20 - ;; -(p8) br.call.sptk.many b6=b6 // ignore this return addr - br.cond.sptk ia64_trace_syscall +#else +(p15) ssm psr.i // M2 restore psr.i +#endif +(p14) br.call.sptk.many b6=b6 // B invoke syscall-handker (ignore return addr) + br.cond.spnt.many ia64_trace_syscall // B do syscall-tracing thingamagic // NOT REACHED +/////////////////////////////////////////////////////////////////////// + // On entry, we optimistically assumed that we're coming from user-space. + // For the rare cases where a system-call is done from within the kernel, + // we fix things up at this point: +.break_fixup: + add r1=-IA64_PT_REGS_SIZE,sp // A allocate space for pt_regs structure + mov ar.rnat=r24 // M2 restore kernel's AR.RNAT + ;; + mov ar.bspstore=r23 // M2 restore kernel's AR.BSPSTORE + br.cond.sptk .back_from_break_fixup END(break_fault) .org ia64_ivt+0x3000 @@ -1201,8 +1273,6 @@ END(interrupt) * - r31: saved pr * - b0: original contents (to be saved) * On exit: - * - executing on bank 1 registers - * - psr.ic enabled, interrupts restored * - p10: TRUE if syscall is invoked with more than 8 out * registers or r15's Nat is true * - r1: kernel's gp @@ -1210,8 +1280,11 @@ END(interrupt) * - r8: -EINVAL if p10 is true * - r12: points to kernel stack * - r13: points to current task + * - r14: preserved (same as on entry) + * - p13: preserved * - p15: TRUE if interrupts need to be re-enabled * - ar.fpsr: set to kernel settings + * - b6: preserved (same as on entry) */ #ifndef CONFIG_XEN GLOBAL_ENTRY(ia64_syscall_setup) @@ -1280,10 +1353,10 @@ GLOBAL_ENTRY(ia64_syscall_setup) (p13) mov in5=-1 ;; st8 [r16]=r21,PT(R8)-PT(AR_FPSR) // save ar.fpsr - tnat.nz p14,p0=in6 + tnat.nz p13,p0=in6 cmp.lt p10,p9=r11,r8 // frame size can't be more than local+8 ;; - stf8 [r16]=f1 // ensure pt_regs.r8 != 0 (see handle_syscall_error) + mov r8=1 (p9) tnat.nz p10,p0=r15 adds r12=-16,r1 // switch to kernel memory stack (with 16 bytes of scratch) @@ -1294,9 +1367,9 @@ GLOBAL_ENTRY(ia64_syscall_setup) mov r13=r2 // establish `current' movl r1=__gp // establish kernel global pointer ;; -(p14) mov in6=-1 + st8 [r16]=r8 // ensure pt_regs.r8 != 0 (see handle_syscall_error) +(p13) mov in6=-1 (p8) mov in7=-1 - nop.i 0 cmp.eq pSys,pNonSys=r0,r0 // set pSys=1, pNonSys=0 movl r17=FPSR_DEFAULT @@ -1323,6 +1396,8 @@ END(ia64_syscall_setup) * element, followed by the arguments. */ ENTRY(dispatch_illegal_op_fault) + .prologue + .body SAVE_MIN_WITH_COVER ssm psr.ic | PSR_DEFAULT_BITS ;; @@ -1335,6 +1410,7 @@ ENTRY(dispatch_illegal_op_fault) mov out0=ar.ec ;; SAVE_REST + PT_REGS_UNWIND_INFO(0) ;; br.call.sptk.many rp=ia64_illegal_op_fault .ret0: ;; @@ -1365,6 +1441,8 @@ END(dispatch_illegal_op_fault) FAULT(17) ENTRY(non_syscall) + mov ar.rsc=r27 // restore ar.rsc before SAVE_MIN_WITH_COVER + ;; SAVE_MIN_WITH_COVER // There is no particular reason for this code to be here, other than that @@ -1540,7 +1618,7 @@ ENTRY(daccess_rights) ;; ld8 r16=[r16] ;; - XEN_HYPER_RSM_PSR_DT; + XEN_HYPER_RSM_PSR_DT #else mov r16=cr.ifa rsm psr.dt @@ -1584,6 +1662,25 @@ END(disabled_fp_reg) // 0x5600 Entry 26 (size 16 bundles) Nat Consumption (11,23,37,50) ENTRY(nat_consumption) DBG_FAULT(26) + + mov r16=cr.ipsr + mov r17=cr.isr + mov r31=pr // save PR + ;; + and r18=0xf,r17 // r18 = cr.ipsr.code{3:0} + tbit.z p6,p0=r17,IA64_ISR_NA_BIT + ;; + cmp.ne.or p6,p0=IA64_ISR_CODE_LFETCH,r18 + dep r16=-1,r16,IA64_PSR_ED_BIT,1 +(p6) br.cond.spnt 1f // branch if (cr.ispr.na == 0 || cr.ipsr.code{3:0} != LFETCH) + ;; + mov cr.ipsr=r16 // set cr.ipsr.na + mov pr=r31,-1 + ;; + rfi + +1: mov pr=r31,-1 + ;; FAULT(26) END(nat_consumption) @@ -1624,7 +1721,7 @@ ENTRY(speculation_vector) #ifdef CONFIG_XEN XEN_HYPER_RFI; #else - rfi + rfi // and go back #endif END(speculation_vector) @@ -1647,7 +1744,6 @@ END(debug_vector) // 0x5a00 Entry 30 (size 16 bundles) Unaligned Reference (57) ENTRY(unaligned_access) DBG_FAULT(30) - mov r16=cr.ipsr mov r31=pr // prepare to save predicates ;; br.sptk.many dispatch_unaligned_handler _______________________________________________ Xen-changelog mailing list Xen-changelog@xxxxxxxxxxxxxxxxxxx http://lists.xensource.com/xen-changelog
|
Lists.xenproject.org is hosted with RackSpace, monitoring our |