[Qemu-devel] [PATCH] linux-user: fix segmentation fault passing with g2h(x) != x

[Alexander Graf]

When forwarding a segmentation fault into the guest process, we were passing
the host's address directly into the guest process's signal descriptor.

That obviously confused the guest process, since it didn't know what to make
of the (usually 32-bit truncated) address. Passing in g2h(address) makes the
guest process a lot happier.

This fixes java running in arm-linux-user for me.

Signed-off-by: Alexander Graf <address@hidden>
---
 user-exec.c |   25 +++++++++++++------------
 1 files changed, 13 insertions(+), 12 deletions(-)

diff --git a/user-exec.c b/user-exec.c
index 36d29b4..83d2d44 100644
--- a/user-exec.c
+++ b/user-exec.c
@@ -100,19 +100,20 @@ static inline int handle_cpu_signal(uintptr_t pc, 
unsigned long address,
     /* Maybe we're still holding the TB fiddling lock? */
     spin_unlock_safe(&tb_lock);
 
-    /* XXX: locking issue */
-    if (is_write && h2g_valid(address)
-        && page_unprotect(h2g(address), pc, puc)) {
-        return 1;
-    }
+    if (h2g_valid(address)) {
+        /* XXX: locking issue */
+        if (is_write && page_unprotect(h2g(address), pc, puc)) {
+            return 1;
+        }
 
-    /* see if it is an MMU fault */
-    ret = cpu_handle_mmu_fault(env, address, is_write, MMU_USER_IDX);
-    if (ret < 0) {
-        return 0; /* not an MMU fault */
-    }
-    if (ret == 0) {
-        return 1; /* the MMU fault was handled without causing real CPU fault 
*/
+        /* see if it is an MMU fault */
+        ret = cpu_handle_mmu_fault(env, h2g(address), is_write, MMU_USER_IDX);
+        if (ret < 0) {
+            return 0; /* not an MMU fault */
+        }
+        if (ret == 0) {
+            return 1; /* the MMU fault was handled without causing real CPU 
fault */
+        }
     }
     /* now we have a real cpu fault */
     tb = tb_find_pc(pc);
-- 
1.6.0.2