qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Ani Sinha <anisinha@redhat.com>
To: "Cleber Rosa" <crosa@redhat.com>,
	"Philippe Mathieu-Daudé" <philmd@linaro.org>,
	"Wainer dos Santos Moschetta" <wainersm@redhat.com>,
	"Beraldo Leal" <bleal@redhat.com>
Cc: Ani Sinha <anisinha@redhat.com>,
	imammedo@redhat.com, David Hildenbrand <david@redhat.com>,
	qemu-devel@nongnu.org
Subject: [PATCH] tests/avocado: add test to exercize processor address space memory bound checks
Date: Fri,  6 Oct 2023 13:00:22 +0530	[thread overview]
Message-ID: <20231006073022.120279-1-anisinha@redhat.com> (raw)

QEMU has validations to make sure that a VM is not started with more memory
(static and hotpluggable memory) than what the guest processor can address
directly with its addressing bits. This change adds a test to make sure QEMU
fails to start with a specific error message when an attempt is made to
start a VM with more memory than what the processor can directly address.
The test also checks for passing cases when the address space of the processor
is capable of addressing all memory. Boundary cases are tested.

CC: imammedo@redhat.com
CC: David Hildenbrand <david@redhat.com>
Signed-off-by: Ani Sinha <anisinha@redhat.com>
---
 tests/avocado/mem-addr-space-check.py | 113 ++++++++++++++++++++++++++
 1 file changed, 113 insertions(+)
 create mode 100644 tests/avocado/mem-addr-space-check.py

diff --git a/tests/avocado/mem-addr-space-check.py b/tests/avocado/mem-addr-space-check.py
new file mode 100644
index 0000000000..6949f7567e
--- /dev/null
+++ b/tests/avocado/mem-addr-space-check.py
@@ -0,0 +1,113 @@
+# Check for crash when using memory beyond the available guest processor
+# address space.
+#
+# Copyright (c) 2023 Red Hat, Inc.
+#
+# Author:
+#  Ani Sinha <anisinha@redhat.com>
+#
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 2 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+from avocado_qemu import QemuSystemTest
+import signal
+
+class MemAddrCheck(QemuSystemTest):
+    def test_phybits_low_pse36(self):
+        """
+        :avocado: tags=machine:q35
+        :avocado: tags=arch:x86_64
+
+        With pse36 feature ON, a processor has 36 bits of addressing. So it can
+        access up to a maximum of 64GiB of memory. Memory hotplug region begins
+        at 4 GiB boundary when "above_4g_mem_size" is 0 (this would be true when
+        we have 0.5 GiB of VM memory, see pc_q35_init()). This means total
+        hotpluggable memory size is 60 GiB. Per slot, we reserve 1 GiB of memory
+        for dimm alignment for all newer machines (see enforce_aligned_dimm
+        property for pc machines and pc_get_device_memory_range()). That leaves
+        total hotpluggable actual memory size of 59 GiB. If the VM is started
+        with 0.5 GiB of memory, maxmem should be set to a maximum value of
+        59.5 GiB to ensure that the processor can address all memory directly.
+        If maxmem is set to 59.6G, QEMU should fail to start with a message that
+        says "phy-bits are too low".
+        If maxmem is set to 59.5G with all other QEMU parameters identical, QEMU
+        should start fine.
+        """
+        self.vm.add_args('-S', '-machine', 'q35', '-m',
+                         '512,slots=1,maxmem=59.6G',
+                         '-cpu', 'pentium,pse36=on', '-display', 'none',
+                         '-object', 'memory-backend-ram,id=mem1,size=1G',
+                         '-device', 'virtio-mem-pci,id=vm0,memdev=mem1')
+        self.vm.set_qmp_monitor(enabled=False)
+        self.vm.launch()
+        self.vm.wait()
+        self.assertEquals(self.vm.exitcode(), 1, "QEMU exit code should be 1")
+        self.assertRegex(self.vm.get_log(), r'phys-bits too low')
+
+    def test_phybits_ok_pentium_pse36(self):
+        """
+        :avocado: tags=machine:q35
+        :avocado: tags=arch:x86_64
+
+        Setting maxmem to 59.5G and making sure that QEMU can start with the
+        same options as the failing case above.
+        """
+        self.vm.add_args('-machine', 'q35', '-m',
+                         '512,slots=1,maxmem=59.5G',
+                         '-cpu', 'pentium,pse36=on', '-display', 'none',
+                         '-object', 'memory-backend-ram,id=mem1,size=1G',
+                         '-device', 'virtio-mem-pci,id=vm0,memdev=mem1')
+        self.vm.set_qmp_monitor(enabled=False)
+        self.vm.launch()
+        self.vm.shutdown()
+        self.assertEquals(self.vm.exitcode(), -signal.SIGTERM,
+                          "QEMU did not terminate gracefully upon SIGTERM")
+
+    def test_phybits_ok_pentium2(self):
+        """
+        :avocado: tags=machine:q35
+        :avocado: tags=arch:x86_64
+
+        Pentium2 has 36 bits of addressing, so its same as pentium
+        with pse36 ON.
+        """
+        self.vm.add_args('-machine', 'q35', '-m',
+                         '512,slots=1,maxmem=59.5G',
+                         '-cpu', 'pentium2', '-display', 'none',
+                         '-object', 'memory-backend-ram,id=mem1,size=1G',
+                         '-device', 'virtio-mem-pci,id=vm0,memdev=mem1')
+        self.vm.set_qmp_monitor(enabled=False)
+        self.vm.launch()
+        self.vm.shutdown()
+        self.assertEquals(self.vm.exitcode(), -signal.SIGTERM,
+                          "QEMU did not terminate gracefully upon SIGTERM")
+
+    def test_phybits_low_nonpse36(self):
+        """
+        :avocado: tags=machine:q35
+        :avocado: tags=arch:x86_64
+
+        Pentium processor has 32 bits of addressing without pse36 or pae
+        so it can access up to 4 GiB of memory. Setting maxmem to 4GiB
+        should make QEMU fail to start with "phys-bits too low" message.
+        """
+        self.vm.add_args('-S', '-machine', 'q35', '-m',
+                         '512,slots=1,maxmem=4G',
+                         '-cpu', 'pentium', '-display', 'none',
+                         '-object', 'memory-backend-ram,id=mem1,size=1G',
+                         '-device', 'virtio-mem-pci,id=vm0,memdev=mem1')
+        self.vm.set_qmp_monitor(enabled=False)
+        self.vm.launch()
+        self.vm.wait()
+        self.assertEquals(self.vm.exitcode(), 1, "QEMU exit code should be 1")
+        self.assertRegex(self.vm.get_log(), r'phys-bits too low')
-- 
2.42.0



                 reply	other threads:[~2023-10-06  7:32 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

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=20231006073022.120279-1-anisinha@redhat.com \
    --to=anisinha@redhat.com \
    --cc=bleal@redhat.com \
    --cc=crosa@redhat.com \
    --cc=david@redhat.com \
    --cc=imammedo@redhat.com \
    --cc=philmd@linaro.org \
    --cc=qemu-devel@nongnu.org \
    --cc=wainersm@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).