[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Qemu-devel] [PATCH 43/48] mmu-hash*: Don't use full ppc_hash{32, 64}_tr
From: |
David Gibson |
Subject: |
[Qemu-devel] [PATCH 43/48] mmu-hash*: Don't use full ppc_hash{32, 64}_translate() path for get_phys_page_debug() |
Date: |
Tue, 12 Mar 2013 21:31:45 +1100 |
Currently the hash mmu versionsof get_phys_page_debug() use the same
ppc64_hash64_translate() function to do the translation logic as the normal
mm fault handler code.
That sounds like a good idea, but has some complications. The debug path
doesn't need, or even want some parts of the full translation path, like
permissions checking. Furthermore, the pte flags update included in the
normal path means that the debug call is not quite side effect free.
This patch, therefore, reimplements get_phys_page_debug as the minimal
required subset of the full translation path.
Signed-off-by: David Gibson <address@hidden>`z
---
target-ppc/mmu-hash32.c | 34 +++++++++++++++++++++++++++-------
target-ppc/mmu-hash64.c | 19 ++++++++++++++++---
2 files changed, 43 insertions(+), 10 deletions(-)
diff --git a/target-ppc/mmu-hash32.c b/target-ppc/mmu-hash32.c
index 07e9b8c..5ec1a09 100644
--- a/target-ppc/mmu-hash32.c
+++ b/target-ppc/mmu-hash32.c
@@ -437,18 +437,38 @@ static int ppc_hash32_translate(CPUPPCState *env, struct
mmu_ctx_hash32 *ctx,
return 0;
}
-hwaddr ppc_hash32_get_phys_page_debug(CPUPPCState *env, target_ulong addr)
+hwaddr ppc_hash32_get_phys_page_debug(CPUPPCState *env, target_ulong eaddr)
{
- struct mmu_ctx_hash32 ctx;
+ target_ulong sr;
+ hwaddr pte_offset;
+ ppc_hash_pte32_t pte;
+ int prot;
+
+ if (msr_dr == 0) {
+ /* Translation is off */
+ return eaddr;
+ }
+
+ if (env->nb_BATs != 0) {
+ hwaddr raddr = ppc_hash32_bat_lookup(env, eaddr, 0, &prot);
+ if (raddr != -1) {
+ return raddr;
+ }
+ }
+
+ sr = env->sr[eaddr >> 28];
+
+ if (sr & SR32_T) {
+ /* FIXME: Add suitable debug support for Direct Store segments */
+ return -1;
+ }
- /* FIXME: Will not behave sanely for direct store segments, but
- * they're almost never used */
- if (unlikely(ppc_hash32_translate(env, &ctx, addr, 0)
- != 0)) {
+ pte_offset = ppc_hash32_htab_lookup(env, sr, eaddr, &pte);
+ if (pte_offset == -1) {
return -1;
}
- return ctx.raddr & TARGET_PAGE_MASK;
+ return ppc_hash32_pte_raddr(sr, pte, eaddr) & TARGET_PAGE_MASK;
}
int ppc_hash32_handle_mmu_fault(CPUPPCState *env, target_ulong address, int
rwx,
diff --git a/target-ppc/mmu-hash64.c b/target-ppc/mmu-hash64.c
index d986c0f..4a7dbbb 100644
--- a/target-ppc/mmu-hash64.c
+++ b/target-ppc/mmu-hash64.c
@@ -449,13 +449,26 @@ static int ppc_hash64_translate(CPUPPCState *env, struct
mmu_ctx_hash64 *ctx,
hwaddr ppc_hash64_get_phys_page_debug(CPUPPCState *env, target_ulong addr)
{
- struct mmu_ctx_hash64 ctx;
+ ppc_slb_t *slb;
+ hwaddr pte_offset;
+ ppc_hash_pte64_t pte;
+
+ if (msr_dr == 0) {
+ /* In real mode the top 4 effective address bits are ignored */
+ return addr & 0x0FFFFFFFFFFFFFFFULL;
+ }
- if (unlikely(ppc_hash64_translate(env, &ctx, addr, 0) != 0)) {
+ slb = slb_lookup(env, addr);
+ if (!slb) {
+ return -1;
+ }
+
+ pte_offset = ppc_hash64_htab_lookup(env, slb, addr, &pte);
+ if (pte_offset == -1) {
return -1;
}
- return ctx.raddr & TARGET_PAGE_MASK;
+ return ppc_hash64_pte_raddr(slb, pte, addr) & TARGET_PAGE_MASK;
}
int ppc_hash64_handle_mmu_fault(CPUPPCState *env, target_ulong address, int
rwx,
--
1.7.10.4
- [Qemu-devel] [PATCH 28/48] mmu-hash*: Remove permission checking from find_pte{32, 64}(), (continued)
- [Qemu-devel] [PATCH 28/48] mmu-hash*: Remove permission checking from find_pte{32, 64}(), David Gibson, 2013/03/12
- [Qemu-devel] [PATCH 29/48] mmu-hash64: Clean up ppc_hash64_htab_lookup(), David Gibson, 2013/03/12
- [Qemu-devel] [PATCH 41/48] mmu-hash*: Clean up real address calculation, David Gibson, 2013/03/12
- [Qemu-devel] [PATCH 48/48] target-ppc: Use QOM method dispatch for MMU fault handling, David Gibson, 2013/03/12
- [Qemu-devel] [PATCH 45/48] mmu-hash64: Implement Virtual Page Class Key Protection, David Gibson, 2013/03/12
- [Qemu-devel] [PATCH 40/48] mmu-hash*: Clean up PTE flags update, David Gibson, 2013/03/12
- [Qemu-devel] [PATCH 47/48] target-ppc: Move ppc tlb_fill implementation into mmu_helper.c, David Gibson, 2013/03/12
- [Qemu-devel] [PATCH 44/48] mmu-hash*: Merge translate and fault handling functions, David Gibson, 2013/03/12
- [Qemu-devel] [PATCH 46/48] target-ppc: Split user only code out of mmu_helper.c, David Gibson, 2013/03/12
- [Qemu-devel] [PATCH 43/48] mmu-hash*: Don't use full ppc_hash{32, 64}_translate() path for get_phys_page_debug(),
David Gibson <=
- Re: [Qemu-devel] [0/48] target-ppc: MMU implementation cleanup for hash MMUs, Alexander Graf, 2013/03/21