From: Michael Roth <mdroth@linux.vnet.ibm.com>
To: qemu-devel@nongnu.org
Cc: Max Filippov <jcmvbkbc@gmail.com>, qemu-stable@nongnu.org
Subject: [Qemu-devel] [PATCH 20/43] target-xtensa: fix translation for opcodes crossing page boundary
Date: Tue, 24 Feb 2015 15:47:55 -0600 [thread overview]
Message-ID: <1424814498-6993-21-git-send-email-mdroth@linux.vnet.ibm.com> (raw)
In-Reply-To: <1424814498-6993-1-git-send-email-mdroth@linux.vnet.ibm.com>
From: Max Filippov <jcmvbkbc@gmail.com>
If TB ends with an opcode that crosses page boundary and the following
page is not executable then EPC1 for the code fetch exception wrongly
points at the beginning of the TB. Always treat instruction that crosses
page boundary as a separate TB.
Cc: qemu-stable@nongnu.org
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
(cherry picked from commit 01673a3401614b4199c9946ad47b97bedfc7a7c2)
Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com>
---
target-xtensa/translate.c | 27 +++++++++++++++++++++++----
1 file changed, 23 insertions(+), 4 deletions(-)
diff --git a/target-xtensa/translate.c b/target-xtensa/translate.c
index badca19..a81573d 100644
--- a/target-xtensa/translate.c
+++ b/target-xtensa/translate.c
@@ -887,6 +887,11 @@ static TCGv_i32 gen_mac16_m(TCGv_i32 v, bool hi, bool is_unsigned)
return m;
}
+static inline unsigned xtensa_op0_insn_len(unsigned op0)
+{
+ return op0 >= 8 ? 2 : 3;
+}
+
static void disas_xtensa_insn(CPUXtensaState *env, DisasContext *dc)
{
#define HAS_OPTION_BITS(opt) do { \
@@ -989,6 +994,7 @@ static void disas_xtensa_insn(CPUXtensaState *env, DisasContext *dc)
uint8_t b0 = cpu_ldub_code(env, dc->pc);
uint8_t b1 = cpu_ldub_code(env, dc->pc + 1);
uint8_t b2 = 0;
+ unsigned len = xtensa_op0_insn_len(OP0);
static const uint32_t B4CONST[] = {
0xffffffff, 1, 2, 3, 4, 5, 6, 7, 8, 10, 12, 16, 32, 64, 128, 256
@@ -998,13 +1004,19 @@ static void disas_xtensa_insn(CPUXtensaState *env, DisasContext *dc)
32768, 65536, 2, 3, 4, 5, 6, 7, 8, 10, 12, 16, 32, 64, 128, 256
};
- if (OP0 >= 8) {
- dc->next_pc = dc->pc + 2;
+ switch (len) {
+ case 2:
HAS_OPTION(XTENSA_OPTION_CODE_DENSITY);
- } else {
- dc->next_pc = dc->pc + 3;
+ break;
+
+ case 3:
b2 = cpu_ldub_code(env, dc->pc + 2);
+ break;
+
+ default:
+ RESERVED();
}
+ dc->next_pc = dc->pc + len;
switch (OP0) {
case 0: /*QRST*/
@@ -2949,6 +2961,12 @@ invalid_opcode:
#undef HAS_OPTION
}
+static inline unsigned xtensa_insn_len(CPUXtensaState *env, DisasContext *dc)
+{
+ uint8_t b0 = cpu_ldub_code(env, dc->pc);
+ return xtensa_op0_insn_len(OP0);
+}
+
static void check_breakpoint(CPUXtensaState *env, DisasContext *dc)
{
CPUState *cs = CPU(xtensa_env_get_cpu(env));
@@ -3081,6 +3099,7 @@ void gen_intermediate_code_internal(XtensaCPU *cpu,
} while (dc.is_jmp == DISAS_NEXT &&
insn_count < max_insns &&
dc.pc < next_page_start &&
+ dc.pc + xtensa_insn_len(env, &dc) <= next_page_start &&
tcg_ctx.gen_opc_ptr < gen_opc_end);
reset_litbase(&dc);
--
1.9.1
next prev parent reply other threads:[~2015-02-24 21:52 UTC|newest]
Thread overview: 50+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-02-24 21:47 [Qemu-devel] Patch Round-up for stable 2.2.1, freeze on 2015-03-05 Michael Roth
2015-02-24 21:47 ` [Qemu-devel] [PATCH 01/43] block: Make essential BlockDriver objects public Michael Roth
2015-02-24 21:47 ` [Qemu-devel] [PATCH 02/43] block: Omit bdrv_find_format for essential drivers Michael Roth
2015-02-24 21:47 ` [Qemu-devel] [PATCH 03/43] block/vvfat: qcow driver may not be found Michael Roth
2015-02-24 21:47 ` [Qemu-devel] [PATCH 04/43] block/nfs: Add create_opts Michael Roth
2015-02-24 21:47 ` [Qemu-devel] [PATCH 05/43] block: Check create_opts before image creation Michael Roth
2015-02-24 21:47 ` [Qemu-devel] [PATCH 06/43] qemu-img: " Michael Roth
2015-02-24 21:47 ` [Qemu-devel] [PATCH 07/43] qemu-img: Check create_opts before image amendment Michael Roth
2015-02-24 21:47 ` [Qemu-devel] [PATCH 08/43] iotests: Only kill NBD server if it runs Michael Roth
2015-02-24 21:47 ` [Qemu-devel] [PATCH 09/43] iotests: Add test for unsupported image creation Michael Roth
2015-02-24 21:47 ` [Qemu-devel] [PATCH 10/43] qcow2: Prevent numerical overflow Michael Roth
2015-02-24 21:47 ` [Qemu-devel] [PATCH 11/43] qcow2: Flushing the caches in qcow2_close may fail Michael Roth
2015-02-24 21:47 ` [Qemu-devel] [PATCH 12/43] qcow2: Respect bdrv_truncate() error Michael Roth
2015-02-24 21:47 ` [Qemu-devel] [PATCH 13/43] block/raw-posix: Fix ret in raw_open_common() Michael Roth
2015-02-24 21:47 ` [Qemu-devel] [PATCH 14/43] block migration: fix return value Michael Roth
2015-02-24 21:47 ` [Qemu-devel] [PATCH 15/43] qcow2: Fix header extension size check Michael Roth
2015-02-24 21:47 ` [Qemu-devel] [PATCH 16/43] qcow2.py: Add required padding for header extensions Michael Roth
2015-02-24 21:47 ` [Qemu-devel] [PATCH 17/43] block: Don't probe for unknown backing file format Michael Roth
2015-02-24 21:47 ` [Qemu-devel] [PATCH 18/43] linuxboot: fix loading old kernels Michael Roth
2015-02-24 21:47 ` [Qemu-devel] [PATCH 19/43] audio: Don't free hw resources until after hw backend is stopped Michael Roth
2015-02-24 21:47 ` Michael Roth [this message]
2015-02-24 21:47 ` [Qemu-devel] [PATCH 21/43] target-xtensa: test cross-page opcode Michael Roth
2015-02-24 21:47 ` [Qemu-devel] [PATCH 22/43] migration/block: fix pending() return value Michael Roth
2015-02-24 21:47 ` [Qemu-devel] [PATCH 23/43] atomic: fix position of volatile qualifier Michael Roth
2015-02-24 21:47 ` [Qemu-devel] [PATCH 24/43] PPC: Fix crash on spapr_tce_table_finalize() Michael Roth
2015-02-24 21:48 ` [Qemu-devel] [PATCH 25/43] vl.c: fix regression when reading machine type from config file Michael Roth
2015-02-24 21:48 ` [Qemu-devel] [PATCH 26/43] serial: reset thri_pending on IER writes with THRI=0 Michael Roth
2015-02-24 21:48 ` [Qemu-devel] [PATCH 27/43] serial: refine serial_thr_ipending_needed Michael Roth
2015-02-24 21:48 ` [Qemu-devel] [PATCH 28/43] pckbd: set bits 2-3-6-7 of the output port by default Michael Roth
2015-02-24 21:48 ` [Qemu-devel] [PATCH 29/43] linux-user: Fix broken m68k signal handling on 64 bit hosts Michael Roth
2015-02-25 8:39 ` Laurent Vivier
2015-02-25 11:28 ` Peter Maydell
2015-02-25 11:58 ` Laurent Vivier
2015-02-25 12:14 ` Peter Maydell
2015-02-24 21:48 ` [Qemu-devel] [PATCH 30/43] scsi: fix cancellation when I/O was completed but DMA was not Michael Roth
2015-02-24 21:48 ` [Qemu-devel] [PATCH 31/43] target-i386: fix movntsd on big-endian hosts Michael Roth
2015-02-24 21:48 ` [Qemu-devel] [PATCH 32/43] vt82c686: avoid out-of-bounds read Michael Roth
2015-02-24 21:48 ` [Qemu-devel] [PATCH 33/43] virtio: fix feature bit checks Michael Roth
2015-02-24 21:48 ` [Qemu-devel] [PATCH 34/43] sb16: fix interrupt acknowledgement Michael Roth
2015-02-24 21:48 ` [Qemu-devel] [PATCH 35/43] hw/input/hid.c Fix capslock hid code Michael Roth
2015-02-24 21:48 ` [Qemu-devel] [PATCH 36/43] target-arm/translate-a64: Fix wrong mmu_idx usage for LDT/STT Michael Roth
2015-02-24 21:48 ` [Qemu-devel] [PATCH 37/43] vfio-pci: Fix missing unparent of dynamically allocated MemoryRegion Michael Roth
2015-02-24 21:48 ` [Qemu-devel] [PATCH 38/43] qemu-thread: fix qemu_event without futexes Michael Roth
2015-02-24 21:48 ` [Qemu-devel] [PATCH 39/43] libcacard: stop linking against every single 3rd party library Michael Roth
2015-02-24 21:48 ` [Qemu-devel] [PATCH 40/43] fix mc146818rtc wrong subsection name to avoid vmstate_subsection_load() fail Michael Roth
2015-02-24 21:48 ` [Qemu-devel] [PATCH 41/43] block/iscsi: fix uninitialized variable Michael Roth
2015-02-24 21:48 ` [Qemu-devel] [PATCH 42/43] qtest: Fix deadloop by running main loop AIO context's timers Michael Roth
2015-02-24 21:48 ` [Qemu-devel] [PATCH 43/43] exec: change default exception_index value for migration to -1 Michael Roth
2015-02-25 2:51 ` [Qemu-devel] [Qemu-stable] Patch Round-up for stable 2.2.1, freeze on 2015-03-05 Gonglei (Arei)
2015-02-25 9:33 ` [Qemu-devel] " Leon Alrae
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=1424814498-6993-21-git-send-email-mdroth@linux.vnet.ibm.com \
--to=mdroth@linux.vnet.ibm.com \
--cc=jcmvbkbc@gmail.com \
--cc=qemu-devel@nongnu.org \
--cc=qemu-stable@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).