lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite for Android: free password hash cracker in your pocket
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20221115012917.1781185-1-debug@rivosinc.com>
Date:   Mon, 14 Nov 2022 17:29:17 -0800
From:   Deepak Gupta <debug@...osinc.com>
To:     debug@...osinc.com
Cc:     aou@...s.berkeley.edu, jan.kiszka@...mens.com, kbingham@...nel.org,
        linux-kernel@...r.kernel.org, linux-riscv@...ts.infradead.org,
        palmer@...belt.com, paul.walmsley@...ive.com
Subject: [PATCH v2] scripts/gdb: add lx_current support for riscv

csr_sscratch CSR holds current task_struct address when hart is in user space.
trap handler on entry spills csr_sscratch into "tp" (x2) register and zeroes out
csr_sscratch CSR. trap handler on exit reloads "tp" with expected user mode value
and place current task_struct address again in csr_scratch CSR.

This patch assumes "tp" is pointing to task_struct. If value in csr_scratch is numerically
greater than "tp" then it assumes csr_scratch is correct address of current task_struct.
This logic holds when
   - hart is in user space, "tp" will be less than csr_scratch.
   - hart is in kernel space but not in trap handler, "tp" will be more than csr_scratch.
   - hart is executing trap handler
           - "tp" is still pointing to user mode but csr_scratch contains ptr to task_struct.
             numerically higher.
           - "tp" is now pointing to task_struct but csr_scratch now contains either 0 or numerically
              smaller value.

Patch also adds new cached type "ulong" in scripts/gdb/linux/utils.py

Signed-off-by: Deepak Gupta <debug@...osinc.com>
---
 scripts/gdb/linux/cpus.py  | 15 +++++++++++++++
 scripts/gdb/linux/utils.py |  5 +++++
 2 files changed, 20 insertions(+)

diff --git a/scripts/gdb/linux/cpus.py b/scripts/gdb/linux/cpus.py
index 15fc4626d236..fd818d7896ce 100644
--- a/scripts/gdb/linux/cpus.py
+++ b/scripts/gdb/linux/cpus.py
@@ -173,6 +173,21 @@ def get_current_task(cpu):
          else:
              raise gdb.GdbError("Sorry, obtaining the current task is not allowed "
                                 "while running in userspace(EL0)")
+    elif utils.is_target_arch("riscv"):
+         current_tp = gdb.parse_and_eval("$tp")
+         scratch_reg = gdb.parse_and_eval("$sscratch")
+
+         # by default tp points to current task
+         current_task = current_tp.cast(task_ptr_type)
+         
+         # scratch register is set 0 in trap handler after entering kernel.
+         # When hart is in user mode, scratch register is pointing to task_struct.
+         # So when scratch register holds higher value (negative address as ulong is bigger value) than tp,
+         # then use scratch register
+         if (scratch_reg.cast(utils.get_ulong_type()) >  current_tp.cast(utils.get_ulong_type())):
+             current_task = scratch_reg.cast(task_ptr_type)
+             
+         return current_task.dereference()
     else:
         raise gdb.GdbError("Sorry, obtaining the current task is not yet "
                            "supported with this arch")
diff --git a/scripts/gdb/linux/utils.py b/scripts/gdb/linux/utils.py
index 1553f68716cc..ddaf3089170d 100644
--- a/scripts/gdb/linux/utils.py
+++ b/scripts/gdb/linux/utils.py
@@ -35,12 +35,17 @@ class CachedType:
 
 
 long_type = CachedType("long")
+ulong_type = CachedType("ulong")
 atomic_long_type = CachedType("atomic_long_t")
 
 def get_long_type():
     global long_type
     return long_type.get_type()
 
+def get_ulong_type():
+    global ulong_type
+    return ulong_type.get_type()
+
 def offset_of(typeobj, field):
     element = gdb.Value(0).cast(typeobj)
     return int(str(element[field].address).split()[0], 16)
-- 
2.25.1

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ