qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Peter Xu <peterx@redhat.com>
To: qemu-devel@nongnu.org, Stefan Hajnoczi <stefanha@redhat.com>
Cc: peterx@redhat.com, Paolo Bonzini <pbonzini@redhat.com>,
	David Hildenbrand <david@redhat.com>
Subject: [PULL 12/14] physmem: teach cpu_memory_rw_debug() to write to more memory regions
Date: Tue, 11 Feb 2025 17:50:56 -0500	[thread overview]
Message-ID: <20250211225059.182533-13-peterx@redhat.com> (raw)
In-Reply-To: <20250211225059.182533-1-peterx@redhat.com>

From: David Hildenbrand <david@redhat.com>

Right now, we only allow for writing to memory regions that allow direct
access using memcpy etc; all other writes are simply ignored. This
implies that debugging guests will not work as expected when writing
to MMIO device regions.

Let's extend cpu_memory_rw_debug() to write to more memory regions,
including MMIO device regions. Reshuffle the condition in
memory_access_is_direct() to make it easier to read and add a comment.

While this change implies that debug access can now also write to MMIO
devices, we now are also permit ELF image loads and similar users of
cpu_memory_rw_debug() to write to MMIO devices; currently we ignore
these writes.

Peter assumes [1] that there's probably a class of guest images, which
will start writing junk (likely zeroes) into device model registers; we
previously would silently ignore any such bogus ELF sections. Likely
these images are of questionable correctness and this can be ignored. If
ever a problem, we could make these cases use address_space_write_rom()
instead, which is left unchanged for now.

This patch is based on previous work by Stefan Zabka.

[1] https://lore.kernel.org/all/CAFEAcA_2CEJKFyjvbwmpt=on=GgMVamQ5hiiVt+zUr6AY3X=Xg@mail.gmail.com/

Resolves: https://gitlab.com/qemu-project/qemu/-/issues/213
Reviewed-by: Peter Xu <peterx@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Link: https://lore.kernel.org/r/20250210084648.33798-8-david@redhat.com
Signed-off-by: Peter Xu <peterx@redhat.com>
---
 include/exec/memattrs.h |  5 ++++-
 include/exec/memory.h   |  3 ++-
 hw/core/cpu-system.c    | 13 +++++++++----
 system/physmem.c        |  9 ++-------
 4 files changed, 17 insertions(+), 13 deletions(-)

diff --git a/include/exec/memattrs.h b/include/exec/memattrs.h
index 060b7e7131..8db1d30464 100644
--- a/include/exec/memattrs.h
+++ b/include/exec/memattrs.h
@@ -44,6 +44,8 @@ typedef struct MemTxAttrs {
      * (see MEMTX_ACCESS_ERROR).
      */
     unsigned int memory:1;
+    /* Debug access that can even write to ROM. */
+    unsigned int debug:1;
     /* Requester ID (for MSI for example) */
     unsigned int requester_id:16;
 
@@ -56,7 +58,8 @@ typedef struct MemTxAttrs {
      * Bus masters which don't specify any attributes will get this
      * (via the MEMTXATTRS_UNSPECIFIED constant), so that we can
      * distinguish "all attributes deliberately clear" from
-     * "didn't specify" if necessary.
+     * "didn't specify" if necessary. "debug" can be set alongside
+     * "unspecified".
      */
     bool unspecified;
 
diff --git a/include/exec/memory.h b/include/exec/memory.h
index b18ecf933e..78c4e0aec8 100644
--- a/include/exec/memory.h
+++ b/include/exec/memory.h
@@ -3018,7 +3018,8 @@ static inline bool memory_access_is_direct(MemoryRegion *mr, bool is_write,
     if (!memory_region_supports_direct_access(mr)) {
         return false;
     }
-    if (is_write) {
+    /* Debug access can write to ROM. */
+    if (is_write && !attrs.debug) {
         return !mr->readonly && !mr->rom_device;
     }
     return true;
diff --git a/hw/core/cpu-system.c b/hw/core/cpu-system.c
index 6aae28a349..6e307c8959 100644
--- a/hw/core/cpu-system.c
+++ b/hw/core/cpu-system.c
@@ -51,13 +51,18 @@ hwaddr cpu_get_phys_page_attrs_debug(CPUState *cpu, vaddr addr,
                                      MemTxAttrs *attrs)
 {
     CPUClass *cc = CPU_GET_CLASS(cpu);
+    hwaddr paddr;
 
     if (cc->sysemu_ops->get_phys_page_attrs_debug) {
-        return cc->sysemu_ops->get_phys_page_attrs_debug(cpu, addr, attrs);
+        paddr = cc->sysemu_ops->get_phys_page_attrs_debug(cpu, addr, attrs);
+    } else {
+        /* Fallback for CPUs which don't implement the _attrs_ hook */
+        *attrs = MEMTXATTRS_UNSPECIFIED;
+        paddr = cc->sysemu_ops->get_phys_page_debug(cpu, addr);
     }
-    /* Fallback for CPUs which don't implement the _attrs_ hook */
-    *attrs = MEMTXATTRS_UNSPECIFIED;
-    return cc->sysemu_ops->get_phys_page_debug(cpu, addr);
+    /* Indicate that this is a debug access. */
+    attrs->debug = 1;
+    return paddr;
 }
 
 hwaddr cpu_get_phys_page_debug(CPUState *cpu, vaddr addr)
diff --git a/system/physmem.c b/system/physmem.c
index 8745c10c9d..d3efdf13d3 100644
--- a/system/physmem.c
+++ b/system/physmem.c
@@ -3680,13 +3680,8 @@ int cpu_memory_rw_debug(CPUState *cpu, vaddr addr,
         if (l > len)
             l = len;
         phys_addr += (addr & ~TARGET_PAGE_MASK);
-        if (is_write) {
-            res = address_space_write_rom(cpu->cpu_ases[asidx].as, phys_addr,
-                                          attrs, buf, l);
-        } else {
-            res = address_space_read(cpu->cpu_ases[asidx].as, phys_addr,
-                                     attrs, buf, l);
-        }
+        res = address_space_rw(cpu->cpu_ases[asidx].as, phys_addr, attrs, buf,
+                               l, is_write);
         if (res != MEMTX_OK) {
             return -1;
         }
-- 
2.47.0



  parent reply	other threads:[~2025-02-11 22:51 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-02-11 22:50 [PULL 00/14] Mem next patches Peter Xu
2025-02-11 22:50 ` [PULL 01/14] system/physmem: take into account fd_offset for file fallocate Peter Xu
2025-02-11 22:50 ` [PULL 02/14] os: add an ability to lock memory on_fault Peter Xu
2025-02-12 14:13   ` Stefan Hajnoczi
2025-02-12 14:17     ` Daniil Tatianin
2025-02-11 22:50 ` [PULL 03/14] system/vl: extract overcommit option parsing into a helper Peter Xu
2025-02-11 22:50 ` [PULL 04/14] system: introduce a new MlockState enum Peter Xu
2025-02-11 22:50 ` [PULL 05/14] overcommit: introduce mem-lock=on-fault Peter Xu
2025-02-11 22:50 ` [PULL 06/14] physmem: factor out memory_region_is_ram_device() check in memory_access_is_direct() Peter Xu
2025-02-11 22:50 ` [PULL 07/14] physmem: factor out RAM/ROMD " Peter Xu
2025-02-11 22:50 ` [PULL 08/14] physmem: factor out direct access check into memory_region_supports_direct_access() Peter Xu
2025-02-11 22:50 ` [PULL 09/14] physmem: disallow direct access to RAM DEVICE in address_space_write_rom() Peter Xu
2025-02-11 22:50 ` [PULL 10/14] memory: pass MemTxAttrs to memory_access_is_direct() Peter Xu
2025-02-11 22:50 ` [PULL 11/14] hmp: use cpu_get_phys_page_debug() in hmp_gva2gpa() Peter Xu
2025-02-11 22:50 ` Peter Xu [this message]
2025-02-11 22:50 ` [PULL 13/14] system/physmem: handle hugetlb correctly in qemu_ram_remap() Peter Xu
2025-02-11 22:50 ` [PULL 14/14] system/physmem: poisoned memory discard on reboot Peter Xu

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=20250211225059.182533-13-peterx@redhat.com \
    --to=peterx@redhat.com \
    --cc=david@redhat.com \
    --cc=pbonzini@redhat.com \
    --cc=qemu-devel@nongnu.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).