Linux CXL
 help / color / mirror / Atom feed
From: Jonathan Cameron <Jonathan.Cameron@huawei.com>
To: qemu-devel@nongnu.org, "Peter Maydell" <peter.maydell@linaro.org>,
	"Gregory Price" <gregory.price@memverge.com>,
	"Alex Bennée" <alex.bennee@linaro.org>,
	richard.henderson@linaro.org
Cc: <linux-cxl@vger.kernel.org>, <linuxarm@huawei.com>
Subject: [RFC PATCH v2] arm/ptw: Handle atomic updates of page tables entries in MMIO during PTW.
Date: Mon, 19 Feb 2024 16:12:29 +0000	[thread overview]
Message-ID: <20240219161229.11776-1-Jonathan.Cameron@huawei.com> (raw)

I'm far from confident this handling here is correct. Hence
RFC.  In particular not sure on what locks I should hold for this
to be even moderately safe.

The function already appears to be inconsistent in what it returns
as the CONFIG_ATOMIC64 block returns the endian converted 'eventual'
value of the cmpxchg whereas the TCG_OVERSIZED_GUEST case returns
the previous value.

Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
---
v2: Thanks Peter for reviewing.
 - Handle the address space as in arm_ldq_ptw() - I should have looked
   at the code immediately above :(
   The result ends up a little more convoluted than I'd like. Could factor
   this block of code out perhaps. I'm also not sure on the fault type
   that is appropriate here.
 - Switch to 'need_lock' as per Philippe's feedback on the x86 fixes.
   likely() doesn't seem appropriate here though.
 
target/arm/ptw.c | 64 ++++++++++++++++++++++++++++++++++++++++++++++--
 1 file changed, 62 insertions(+), 2 deletions(-)

diff --git a/target/arm/ptw.c b/target/arm/ptw.c
index 5eb3577bcd..ba1a27ca2b 100644
--- a/target/arm/ptw.c
+++ b/target/arm/ptw.c
@@ -711,8 +711,68 @@ static uint64_t arm_casq_ptw(CPUARMState *env, uint64_t old_val,
     void *host = ptw->out_host;
 
     if (unlikely(!host)) {
-        fi->type = ARMFault_UnsuppAtomicUpdate;
-        return 0;
+        /* Page table in MMIO Memory Region */
+        CPUState *cs = env_cpu(env);
+        MemTxAttrs attrs = {
+            .space = ptw->out_space,
+            .secure = arm_space_is_secure(ptw->out_space),
+        };
+        AddressSpace *as = arm_addressspace(cs, attrs);
+        MemTxResult result = MEMTX_OK;
+        bool need_lock = !bql_locked();
+
+        if (need_lock) {
+            bql_lock();
+        }
+        if (ptw->out_be) {
+            cur_val = address_space_ldq_be(as, ptw->out_phys, attrs, &result);
+            if (unlikely(result != MEMTX_OK)) {
+                fi->type = ARMFault_SyncExternalOnWalk;
+                fi->ea = arm_extabort_type(result);
+                if (need_lock) {
+                    bql_unlock();
+                }
+                return old_val;
+            }
+            if (cur_val == old_val) {
+                address_space_stq_be(as, ptw->out_phys, new_val, attrs, &result);
+                if (unlikely(result != MEMTX_OK)) {
+                    fi->type = ARMFault_SyncExternalOnWalk;
+                    fi->ea = arm_extabort_type(result);
+                    if (need_lock) {
+                        bql_unlock();
+                    }
+                    return old_val;
+                }
+                cur_val = new_val;
+            }
+        } else {
+            cur_val = address_space_ldq_le(as, ptw->out_phys, attrs, &result);
+            if (unlikely(result != MEMTX_OK)) {
+                fi->type = ARMFault_SyncExternalOnWalk;
+                fi->ea = arm_extabort_type(result);
+                if (need_lock) {
+                    bql_unlock();
+                }
+                return old_val;
+            }
+            if (cur_val == old_val) {
+                address_space_stq_le(as, ptw->out_phys, new_val, attrs, &result);
+                if (unlikely(result != MEMTX_OK)) {
+                    fi->type = ARMFault_SyncExternalOnWalk;
+                    fi->ea = arm_extabort_type(result);
+                    if (need_lock) {
+                        bql_unlock();
+                    }
+                    return old_val;
+                }
+                cur_val = new_val;
+            }
+        }
+        if (need_lock) {
+            bql_unlock();
+        }
+        return cur_val;
     }
 
     /*
-- 
2.39.2


             reply	other threads:[~2024-02-19 16:12 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-02-19 16:12 Jonathan Cameron [this message]
2024-02-22 21:21 ` [RFC PATCH v2] arm/ptw: Handle atomic updates of page tables entries in MMIO during PTW Richard Henderson
2024-02-23 10:02   ` Peter Maydell
2024-02-23 18:08     ` Peter Maydell

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=20240219161229.11776-1-Jonathan.Cameron@huawei.com \
    --to=jonathan.cameron@huawei.com \
    --cc=alex.bennee@linaro.org \
    --cc=gregory.price@memverge.com \
    --cc=linux-cxl@vger.kernel.org \
    --cc=linuxarm@huawei.com \
    --cc=peter.maydell@linaro.org \
    --cc=qemu-devel@nongnu.org \
    --cc=richard.henderson@linaro.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