[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Qemu-ppc] [PATCH 05/10] ppc: Fix generation if ISI/DSI vs. HV mode
From: |
David Gibson |
Subject: |
Re: [Qemu-ppc] [PATCH 05/10] ppc: Fix generation if ISI/DSI vs. HV mode |
Date: |
Wed, 15 Jun 2016 11:09:48 +1000 |
User-agent: |
Mutt/1.6.1 (2016-04-27) |
On Tue, Jun 14, 2016 at 08:42:26AM +0200, Cédric Le Goater wrote:
> On 06/14/2016 08:34 AM, David Gibson wrote:
> > On Mon, Jun 13, 2016 at 07:24:51AM +0200, Cédric Le Goater wrote:
> >> From: Benjamin Herrenschmidt <address@hidden>
> >>
> >> Under some circumstances, we need to direct ISI and DSI interrupts
> >> at the hypervisor, turning them into HISI/HDSI, and using different
> >> SPRs (HDSISR and HDAR) depending on the combination of MSR_DR and
> >> the corresponding VPM bits in LPCR.
> >>
> >> This moves part of the code into helpers that are fixed to select
> >> the right exception type and registers. On pre-P7 processors, LPCR
> >> is 0 which provides the old behaviour of directing the interrupts
> >> at the supervisor.
> >>
> >> Thanks to Andrei Warkentin for finding a bug when HV=1
> >>
> >> Signed-off-by: Benjamin Herrenschmidt <address@hidden>
> >> Reviewed-by: David Gibson <address@hidden>
> >> ---
> >> target-ppc/mmu-hash64.c | 69
> >> +++++++++++++++++++++++++++++++++++--------------
> >> 1 file changed, 50 insertions(+), 19 deletions(-)
> >>
> >> diff --git a/target-ppc/mmu-hash64.c b/target-ppc/mmu-hash64.c
> >> index 668da5e22653..072a952c8bd5 100644
> >> --- a/target-ppc/mmu-hash64.c
> >> +++ b/target-ppc/mmu-hash64.c
> >> @@ -613,6 +613,47 @@ unsigned ppc_hash64_hpte_page_shift_noslb(PowerPCCPU
> >> *cpu,
> >> return 0;
> >> }
> >>
> >> +static void ppc_hash64_set_isi(CPUState *cs, CPUPPCState *env,
> >> + uint64_t error_code)
> >> +{
> >> + bool vpm;
> >> +
> >> + if (msr_ir) {
> >> + vpm = !!(env->spr[SPR_LPCR] & LPCR_VPM1);
> >> + } else {
> >> + vpm = !!(env->spr[SPR_LPCR] & LPCR_VPM0);
> >> + }
> >> + if (vpm && !msr_hv) {
> >> + cs->exception_index = POWERPC_EXCP_HISI;
> >
> > In the ISI case, you use HISI if !msr_hv..
> >
> >> + } else {
> >> + cs->exception_index = POWERPC_EXCP_ISI;
> >> + }
> >> + env->error_code = error_code;
> >> +}
> >> +
> >> +static void ppc_hash64_set_dsi(CPUState *cs, CPUPPCState *env, uint64_t
> >> dar,
> >> + uint64_t dsisr)
> >> +{
> >> + bool vpm;
> >> +
> >> + if (msr_dr) {
> >> + vpm = !!(env->spr[SPR_LPCR] & LPCR_VPM1);
> >> + } else {
> >> + vpm = !!(env->spr[SPR_LPCR] & LPCR_VPM0);
> >> + }
> >> + if (vpm && msr_hv) {
> >> + cs->exception_index = POWERPC_EXCP_HDSI;
> >
> > ..but in the DSI case you use HDSI if msr_hv. Is that really right?
>
> No. I forgot to add this patch from Andrei :
>
>
> https://github.com/legoater/qemu/commit/e218fd3ba945bb0f483f5f12bedbb74d897cd5b9
>
> I will send it as a fix.
Given that there are a number of tweaks this series needs, I'd prefer
to see it folded in rather than as a separate patch.
>
> C.
>
> >> + env->spr[SPR_HDAR] = dar;
> >> + env->spr[SPR_HDSISR] = dsisr;
> >> + } else {
> >> + cs->exception_index = POWERPC_EXCP_DSI;
> >> + env->spr[SPR_DAR] = dar;
> >> + env->spr[SPR_DSISR] = dsisr;
> >> + }
> >> + env->error_code = 0;
> >> +}
> >> +
> >> +
> >> int ppc_hash64_handle_mmu_fault(PowerPCCPU *cpu, vaddr eaddr,
> >> int rwx, int mmu_idx)
> >> {
> >> @@ -623,7 +664,7 @@ int ppc_hash64_handle_mmu_fault(PowerPCCPU *cpu, vaddr
> >> eaddr,
> >> hwaddr pte_offset;
> >> ppc_hash_pte64_t pte;
> >> int pp_prot, amr_prot, prot;
> >> - uint64_t new_pte1;
> >> + uint64_t new_pte1, dsisr;
> >> const int need_prot[] = {PAGE_READ, PAGE_WRITE, PAGE_EXEC};
> >> hwaddr raddr;
> >>
> >> @@ -657,26 +698,21 @@ int ppc_hash64_handle_mmu_fault(PowerPCCPU *cpu,
> >> vaddr eaddr,
> >>
> >> /* 3. Check for segment level no-execute violation */
> >> if ((rwx == 2) && (slb->vsid & SLB_VSID_N)) {
> >> - cs->exception_index = POWERPC_EXCP_ISI;
> >> - env->error_code = 0x10000000;
> >> + ppc_hash64_set_isi(cs, env, 0x10000000);
> >> return 1;
> >> }
> >>
> >> /* 4. Locate the PTE in the hash table */
> >> pte_offset = ppc_hash64_htab_lookup(cpu, slb, eaddr, &pte);
> >> if (pte_offset == -1) {
> >> + dsisr = 0x40000000;
> >> if (rwx == 2) {
> >> - cs->exception_index = POWERPC_EXCP_ISI;
> >> - env->error_code = 0x40000000;
> >> + ppc_hash64_set_isi(cs, env, dsisr);
> >> } else {
> >> - cs->exception_index = POWERPC_EXCP_DSI;
> >> - env->error_code = 0;
> >> - env->spr[SPR_DAR] = eaddr;
> >> if (rwx == 1) {
> >> - env->spr[SPR_DSISR] = 0x42000000;
> >> - } else {
> >> - env->spr[SPR_DSISR] = 0x40000000;
> >> + dsisr |= 0x02000000;
> >> }
> >> + ppc_hash64_set_dsi(cs, env, eaddr, dsisr);
> >> }
> >> return 1;
> >> }
> >> @@ -705,14 +741,9 @@ int ppc_hash64_handle_mmu_fault(PowerPCCPU *cpu,
> >> vaddr eaddr,
> >> /* Access right violation */
> >> qemu_log_mask(CPU_LOG_MMU, "PTE access rejected\n");
> >> if (rwx == 2) {
> >> - cs->exception_index = POWERPC_EXCP_ISI;
> >> - env->error_code = 0x08000000;
> >> + ppc_hash64_set_isi(cs, env, 0x08000000);
> >> } else {
> >> - target_ulong dsisr = 0;
> >> -
> >> - cs->exception_index = POWERPC_EXCP_DSI;
> >> - env->error_code = 0;
> >> - env->spr[SPR_DAR] = eaddr;
> >> + dsisr = 0;
> >> if (need_prot[rwx] & ~pp_prot) {
> >> dsisr |= 0x08000000;
> >> }
> >> @@ -722,7 +753,7 @@ int ppc_hash64_handle_mmu_fault(PowerPCCPU *cpu, vaddr
> >> eaddr,
> >> if (need_prot[rwx] & ~amr_prot) {
> >> dsisr |= 0x00200000;
> >> }
> >> - env->spr[SPR_DSISR] = dsisr;
> >> + ppc_hash64_set_dsi(cs, env, eaddr, dsisr);
> >> }
> >> return 1;
> >> }
> >
>
--
David Gibson | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
| _way_ _around_!
http://www.ozlabs.org/~dgibson
signature.asc
Description: PGP signature
- Re: [Qemu-ppc] [PATCH 02/10] ppc: Create cpu_ppc_set_papr() helper (for LPCR), (continued)
[Qemu-ppc] [PATCH 03/10] ppc: Rework POWER7 & POWER8 exception model (part 2), Cédric Le Goater, 2016/06/13
[Qemu-ppc] [PATCH 04/10] ppc: Fix POWER7 and POWER8 exception definitions, Cédric Le Goater, 2016/06/13
[Qemu-ppc] [PATCH 05/10] ppc: Fix generation if ISI/DSI vs. HV mode, Cédric Le Goater, 2016/06/13
[Qemu-ppc] [PATCH 06/10] ppc: Rework generation of priv and inval interrupts, Cédric Le Goater, 2016/06/13
[Qemu-ppc] [PATCH 07/10] ppc: Add real mode CI load/store instructions for P7 and P8, Cédric Le Goater, 2016/06/13
[Qemu-ppc] [PATCH 08/10] ppc: Turn a bunch of booleans from int to bool, Cédric Le Goater, 2016/06/13
[Qemu-ppc] [PATCH 09/10] ppc: Move exception generation code out of line, Cédric Le Goater, 2016/06/13