qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: "Philippe Mathieu-Daudé" <philmd@redhat.com>
To: qemu-devel@nongnu.org
Cc: "Peter Maydell" <peter.maydell@linaro.org>,
	"Michael S. Tsirkin" <mst@redhat.com>,
	"Richard Henderson" <richard.henderson@linaro.org>,
	"Peter Xu" <peterx@redhat.com>, "Bibo Mao" <maobibo@loongson.cn>,
	"Stefan Hajnoczi" <stefanha@redhat.com>,
	"Paolo Bonzini" <pbonzini@redhat.com>,
	"Philippe Mathieu-Daudé" <philmd@redhat.com>
Subject: [RFC PATCH 03/12] exec/memory_ldst: Return MEMTX_UNALIGNED_ERROR for unaligned addresses
Date: Thu, 20 May 2021 13:09:10 +0200	[thread overview]
Message-ID: <20210520110919.2483190-4-philmd@redhat.com> (raw)
In-Reply-To: <20210520110919.2483190-1-philmd@redhat.com>

All address_space internal handlers have the /* warning: addr
must be aligned */ comment, so we don't expect any caller to
pass unaligned addresses.

Now than we added the MemTxAttrs.aligned attribute, callers
might want to pass unaligned addresses. In case they do, be
ready and return MEMTX_UNALIGNED_ERROR.

Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
---
 memory_ldst.c.inc | 49 +++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 49 insertions(+)

diff --git a/memory_ldst.c.inc b/memory_ldst.c.inc
index 84b868f2946..efeb545479e 100644
--- a/memory_ldst.c.inc
+++ b/memory_ldst.c.inc
@@ -32,6 +32,13 @@ static inline uint32_t glue(address_space_ldl_internal, SUFFIX)(ARG1_DECL,
     MemTxResult r;
     bool release_lock = false;
 
+    if (unlikely(!QEMU_IS_ALIGNED(addr, sizeof(uint32_t)))) {
+        if (result) {
+            *result = MEMTX_UNALIGNED_ERROR;
+        }
+        return (uint32_t)-1; /* XXX */
+    }
+
     RCU_READ_LOCK();
     mr = TRANSLATE(addr, &addr1, &l, false, attrs);
     if (l < 4 || !memory_access_is_direct(mr, false)) {
@@ -101,6 +108,13 @@ static inline uint64_t glue(address_space_ldq_internal, SUFFIX)(ARG1_DECL,
     MemTxResult r;
     bool release_lock = false;
 
+    if (unlikely(!QEMU_IS_ALIGNED(addr, sizeof(uint64_t)))) {
+        if (result) {
+            *result = MEMTX_UNALIGNED_ERROR;
+        }
+        return (uint64_t)-1; /* XXX */
+    }
+
     RCU_READ_LOCK();
     mr = TRANSLATE(addr, &addr1, &l, false, attrs);
     if (l < 8 || !memory_access_is_direct(mr, false)) {
@@ -205,6 +219,13 @@ static inline uint16_t glue(address_space_lduw_internal, SUFFIX)(ARG1_DECL,
     MemTxResult r;
     bool release_lock = false;
 
+    if (unlikely(!QEMU_IS_ALIGNED(addr, sizeof(uint16_t)))) {
+        if (result) {
+            *result = MEMTX_UNALIGNED_ERROR;
+        }
+        return (uint16_t)-1; /* XXX */
+    }
+
     RCU_READ_LOCK();
     mr = TRANSLATE(addr, &addr1, &l, false, attrs);
     if (l < 2 || !memory_access_is_direct(mr, false)) {
@@ -275,6 +296,13 @@ void glue(address_space_stl_notdirty, SUFFIX)(ARG1_DECL,
     uint8_t dirty_log_mask;
     bool release_lock = false;
 
+    if (unlikely(!QEMU_IS_ALIGNED(addr, sizeof(uint32_t)))) {
+        if (result) {
+            *result = MEMTX_UNALIGNED_ERROR;
+        }
+        return;
+    }
+
     RCU_READ_LOCK();
     mr = TRANSLATE(addr, &addr1, &l, true, attrs);
     if (l < 4 || !memory_access_is_direct(mr, true)) {
@@ -312,6 +340,13 @@ static inline void glue(address_space_stl_internal, SUFFIX)(ARG1_DECL,
     MemTxResult r;
     bool release_lock = false;
 
+    if (unlikely(!QEMU_IS_ALIGNED(addr, sizeof(uint32_t)))) {
+        if (result) {
+            *result = MEMTX_UNALIGNED_ERROR;
+        }
+        return;
+    }
+
     RCU_READ_LOCK();
     mr = TRANSLATE(addr, &addr1, &l, true, attrs);
     if (l < 4 || !memory_access_is_direct(mr, true)) {
@@ -408,6 +443,13 @@ static inline void glue(address_space_stw_internal, SUFFIX)(ARG1_DECL,
     MemTxResult r;
     bool release_lock = false;
 
+    if (unlikely(!QEMU_IS_ALIGNED(addr, sizeof(uint16_t)))) {
+        if (result) {
+            *result = MEMTX_UNALIGNED_ERROR;
+        }
+        return;
+    }
+
     RCU_READ_LOCK();
     mr = TRANSLATE(addr, &addr1, &l, true, attrs);
     if (l < 2 || !memory_access_is_direct(mr, true)) {
@@ -472,6 +514,13 @@ static void glue(address_space_stq_internal, SUFFIX)(ARG1_DECL,
     MemTxResult r;
     bool release_lock = false;
 
+    if (unlikely(!QEMU_IS_ALIGNED(addr, sizeof(uint64_t)))) {
+        if (result) {
+            *result = MEMTX_UNALIGNED_ERROR;
+        }
+        return;
+    }
+
     RCU_READ_LOCK();
     mr = TRANSLATE(addr, &addr1, &l, true, attrs);
     if (l < 8 || !memory_access_is_direct(mr, true)) {
-- 
2.26.3



  parent reply	other threads:[~2021-05-20 11:11 UTC|newest]

Thread overview: 16+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-05-20 11:09 [RFC PATCH 00/12] exec/memory: Experimental API to catch unaligned accesses Philippe Mathieu-Daudé
2021-05-20 11:09 ` [RFC PATCH 01/12] exec/memory_ldst: Use correct type sizes Philippe Mathieu-Daudé
2021-05-20 11:09 ` [RFC PATCH 02/12] exec/memattrs: Add attribute/error for address alignment Philippe Mathieu-Daudé
2021-06-02 12:39   ` Stefan Hajnoczi
2021-05-20 11:09 ` Philippe Mathieu-Daudé [this message]
2021-05-20 11:09 ` [RFC PATCH 04/12] exec/memory_ldst_cached: Sort declarations Philippe Mathieu-Daudé
2021-05-20 11:09 ` [RFC PATCH 05/12] exec/memory_ldst_cached: Use correct type size Philippe Mathieu-Daudé
2021-05-20 11:09 ` [RFC PATCH 06/12] exec/memory_ldst_cached: Set MemTxResult on success Philippe Mathieu-Daudé
2021-05-20 11:09 ` [RFC PATCH 07/12] exec/memory_ldst_cached: Document aligned addresses are expected Philippe Mathieu-Daudé
2021-05-20 11:09 ` [RFC PATCH 08/12] exec/memory_ldst_cached: Check address alignment if requested Philippe Mathieu-Daudé
2021-05-20 11:09 ` [RFC PATCH 09/12] hw/virtio: Use correct type sizes Philippe Mathieu-Daudé
2021-05-20 11:09 ` [RFC PATCH 10/12] hw/virtio: Extract virtio_lduw_phys_cached_with_attrs() Philippe Mathieu-Daudé
2021-05-20 11:09 ` [RFC PATCH 11/12] hw/virtio: Have vring_avail_flags() return a boolean value Philippe Mathieu-Daudé
2021-05-20 11:09 ` [RFC PATCH 12/12] hw/virtio: Display error if vring flag field is not aligned Philippe Mathieu-Daudé
2021-06-02 12:44   ` Stefan Hajnoczi
2021-06-01  3:32 ` [RFC PATCH 00/12] exec/memory: Experimental API to catch unaligned accesses Philippe Mathieu-Daudé

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=20210520110919.2483190-4-philmd@redhat.com \
    --to=philmd@redhat.com \
    --cc=maobibo@loongson.cn \
    --cc=mst@redhat.com \
    --cc=pbonzini@redhat.com \
    --cc=peter.maydell@linaro.org \
    --cc=peterx@redhat.com \
    --cc=qemu-devel@nongnu.org \
    --cc=richard.henderson@linaro.org \
    --cc=stefanha@redhat.com \
    /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).