From: Paolo Bonzini <pbonzini@redhat.com>
To: qemu-devel@nongnu.org
Cc: agraf@suse.de
Subject: [Qemu-devel] [PATCH 2/3] softmmu: support up to 12 MMU modes
Date: Fri, 20 Feb 2015 13:45:44 +0100 [thread overview]
Message-ID: <1424436345-37924-3-git-send-email-pbonzini@redhat.com> (raw)
In-Reply-To: <1424436345-37924-1-git-send-email-pbonzini@redhat.com>
At 8k per TLB (for 64-bit host or target), 8 or more modes
make the TLBs bigger than 64k, and some RISC TCG backends do
not like that. On the affected hosts, cut the TLB size in
half---there is still a measurable speedup on PPC with the
next patch.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
include/exec/cpu-defs.h | 34 +++++++++++++++-
include/exec/cpu_ldst.h | 104 +++++++++++++++++++++++++++++++++++++++++++++---
2 files changed, 130 insertions(+), 8 deletions(-)
diff --git a/include/exec/cpu-defs.h b/include/exec/cpu-defs.h
index 0ca6f0b..2662de5 100644
--- a/include/exec/cpu-defs.h
+++ b/include/exec/cpu-defs.h
@@ -27,6 +27,7 @@
#include <inttypes.h>
#include "qemu/osdep.h"
#include "qemu/queue.h"
+#include "tcg-target.h"
#ifndef CONFIG_USER_ONLY
#include "exec/hwaddr.h"
#endif
@@ -69,8 +70,6 @@ typedef uint64_t target_ulong;
#define TB_JMP_PAGE_MASK (TB_JMP_CACHE_SIZE - TB_JMP_PAGE_SIZE)
#if !defined(CONFIG_USER_ONLY)
-#define CPU_TLB_BITS 8
-#define CPU_TLB_SIZE (1 << CPU_TLB_BITS)
/* use a fully associative victim tlb of 8 entries */
#define CPU_VTLB_SIZE 8
@@ -80,6 +79,37 @@ typedef uint64_t target_ulong;
#define CPU_TLB_ENTRY_BITS 5
#endif
+/* TCG_TARGET_TLB_DISPLACEMENT_BITS is used in CPU_TLB_BITS to ensure that
+ * the TLB is not unnecessarily small, but still small enough for the
+ * TLB lookup instruction sequence used by the TCG target.
+ *
+ * TCG will have to generate an operand as large as the distance between
+ * tlb_table[0][0] and the tlb_table[NB_MMU_MODES - 1][0].addend. For
+ * simplicity, the TCG targets just round everything up to the next
+ * power of two, and count bits. This works because: 1) the size of each
+ * TLB is a largish power of two, 2) and because the limit of the
+ * displacement is really close to a power of two.
+ *
+ * For example, the maximum displacement 0xFFF0 on PPC and MIPS, but TCG
+ * just says "the displacement is 16 bits". TCG_TARGET_TLB_DISPLACEMENT_BITS
+ * then ensures that tlb_table at least 0x8000 bytes large ("not unnecessarily
+ * small": 2^15). The operand then will come up smaller than 0xFFF0 without
+ * any particular care, because the TLB for a single MMU mode is larger than
+ * 0x10000-0xFFF0=16 bytes. In the end, the maximum value of the operand
+ * will be something like 0xC018; here, 0xC000 is the offset of the last TLB
+ * table (tlb_table[NB_MMU_MODES - 1][0]), and 0x18 is the offset of the
+ * addend field in each TLB entry.
+ */
+#define CPU_TLB_BITS \
+ MIN(8, \
+ TCG_TARGET_TLB_DISPLACEMENT_BITS - CPU_TLB_ENTRY_BITS - \
+ (NB_MMU_MODES <= 1 ? 0 : \
+ NB_MMU_MODES <= 2 ? 1 : \
+ NB_MMU_MODES <= 4 ? 2 : \
+ NB_MMU_MODES <= 8 ? 3 : 4))
+
+#define CPU_TLB_SIZE (1 << CPU_TLB_BITS)
+
typedef struct CPUTLBEntry {
/* bit TARGET_LONG_BITS to TARGET_PAGE_BITS : virtual address
bit TARGET_PAGE_BITS-1..4 : Nonzero for accesses that should not
diff --git a/include/exec/cpu_ldst.h b/include/exec/cpu_ldst.h
index 1673287..0ec398c 100644
--- a/include/exec/cpu_ldst.h
+++ b/include/exec/cpu_ldst.h
@@ -263,12 +263,104 @@ uint64_t helper_ldq_cmmu(CPUArchState *env, target_ulong addr, int mmu_idx);
#undef MEMSUFFIX
#endif /* (NB_MMU_MODES >= 7) */
-#if (NB_MMU_MODES > 7)
-/* Note that supporting NB_MMU_MODES == 9 would require
- * changes to at least the ARM TCG backend.
- */
-#error "NB_MMU_MODES > 7 is not supported for now"
-#endif /* (NB_MMU_MODES > 7) */
+#if (NB_MMU_MODES >= 8) && defined(MMU_MODE7_SUFFIX)
+
+#define CPU_MMU_INDEX 7
+#define MEMSUFFIX MMU_MODE7_SUFFIX
+#define DATA_SIZE 1
+#include "exec/cpu_ldst_template.h"
+
+#define DATA_SIZE 2
+#include "exec/cpu_ldst_template.h"
+
+#define DATA_SIZE 4
+#include "exec/cpu_ldst_template.h"
+
+#define DATA_SIZE 8
+#include "exec/cpu_ldst_template.h"
+#undef CPU_MMU_INDEX
+#undef MEMSUFFIX
+#endif /* (NB_MMU_MODES >= 8) */
+
+#if (NB_MMU_MODES >= 9) && defined(MMU_MODE8_SUFFIX)
+
+#define CPU_MMU_INDEX 8
+#define MEMSUFFIX MMU_MODE8_SUFFIX
+#define DATA_SIZE 1
+#include "exec/cpu_ldst_template.h"
+
+#define DATA_SIZE 2
+#include "exec/cpu_ldst_template.h"
+
+#define DATA_SIZE 4
+#include "exec/cpu_ldst_template.h"
+
+#define DATA_SIZE 8
+#include "exec/cpu_ldst_template.h"
+#undef CPU_MMU_INDEX
+#undef MEMSUFFIX
+#endif /* (NB_MMU_MODES >= 9) */
+
+#if (NB_MMU_MODES >= 10) && defined(MMU_MODE9_SUFFIX)
+
+#define CPU_MMU_INDEX 9
+#define MEMSUFFIX MMU_MODE9_SUFFIX
+#define DATA_SIZE 1
+#include "exec/cpu_ldst_template.h"
+
+#define DATA_SIZE 2
+#include "exec/cpu_ldst_template.h"
+
+#define DATA_SIZE 4
+#include "exec/cpu_ldst_template.h"
+
+#define DATA_SIZE 8
+#include "exec/cpu_ldst_template.h"
+#undef CPU_MMU_INDEX
+#undef MEMSUFFIX
+#endif /* (NB_MMU_MODES >= 10) */
+
+#if (NB_MMU_MODES >= 11) && defined(MMU_MODE10_SUFFIX)
+
+#define CPU_MMU_INDEX 10
+#define MEMSUFFIX MMU_MODE10_SUFFIX
+#define DATA_SIZE 1
+#include "exec/cpu_ldst_template.h"
+
+#define DATA_SIZE 2
+#include "exec/cpu_ldst_template.h"
+
+#define DATA_SIZE 4
+#include "exec/cpu_ldst_template.h"
+
+#define DATA_SIZE 8
+#include "exec/cpu_ldst_template.h"
+#undef CPU_MMU_INDEX
+#undef MEMSUFFIX
+#endif /* (NB_MMU_MODES >= 11) */
+
+#if (NB_MMU_MODES >= 12) && defined(MMU_MODE11_SUFFIX)
+
+#define CPU_MMU_INDEX 11
+#define MEMSUFFIX MMU_MODE11_SUFFIX
+#define DATA_SIZE 1
+#include "exec/cpu_ldst_template.h"
+
+#define DATA_SIZE 2
+#include "exec/cpu_ldst_template.h"
+
+#define DATA_SIZE 4
+#include "exec/cpu_ldst_template.h"
+
+#define DATA_SIZE 8
+#include "exec/cpu_ldst_template.h"
+#undef CPU_MMU_INDEX
+#undef MEMSUFFIX
+#endif /* (NB_MMU_MODES >= 12) */
+
+#if (NB_MMU_MODES > 12)
+#error "NB_MMU_MODES > 12 is not supported for now"
+#endif /* (NB_MMU_MODES > 12) */
/* these access are slower, they must be as rare as possible */
#define CPU_MMU_INDEX (cpu_mmu_index(env))
--
2.3.0
next prev parent reply other threads:[~2015-02-20 12:46 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-02-20 12:45 [Qemu-devel] [PATCH for-2.3 0/3] Support more than 8 MMU modes, speedup by 10% Paolo Bonzini
2015-02-20 12:45 ` [Qemu-devel] [PATCH 1/3] tcg: add TCG_TARGET_TLB_DISPLACEMENT_BITS Paolo Bonzini
2015-02-20 12:45 ` Paolo Bonzini [this message]
2015-02-20 12:45 ` [Qemu-devel] [PATCH 3/3] target-ppc: use separate indices for various translation modes Paolo Bonzini
2015-02-20 13:00 ` Alexander Graf
2015-02-20 17:58 ` Paolo Bonzini
-- strict thread matches above, loose matches on Subject: below --
2015-02-20 17:57 [Qemu-devel] [PATCH v2 for-2.3 0/3] Support more than 8 MMU modes, speedup PPC by 10% Paolo Bonzini
2015-02-20 17:57 ` [Qemu-devel] [PATCH 2/3] softmmu: support up to 12 MMU modes Paolo Bonzini
2015-02-20 19:51 ` Richard Henderson
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1424436345-37924-3-git-send-email-pbonzini@redhat.com \
--to=pbonzini@redhat.com \
--cc=agraf@suse.de \
--cc=qemu-devel@nongnu.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).