[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Qemu-devel] [PATCH RFC 0/2] virt bare-metal payload infrastructure
From: |
Paolo Bonzini |
Subject: |
Re: [Qemu-devel] [PATCH RFC 0/2] virt bare-metal payload infrastructure with atomic test case |
Date: |
Thu, 07 May 2015 14:55:03 +0200 |
User-agent: |
Mozilla/5.0 (X11; Linux x86_64; rv:31.0) Gecko/20100101 Thunderbird/31.5.0 |
On 07/05/2015 13:31, Alexander Spyridakis wrote:
> This series implements a bare-metal test payload for the ARM and AARCH64 virt
> machine models. With the new direction of TCG to finally get multi-threaded
> capabilities, simple and easily deployed tests are needed, to reproduce race
> conditions and intuitively debug QEMU's TCG internals.
>
> The goal of the series is to provide an easy way to create SMP guest test
> cases
> with minimal initialization. In its current state the provided features are,
> SMP functionality through PSCI calls, a simple spinlock test-case, and
> a minimal printf implementation based on the multiboot test. Parts of this
> payload have been also tested with normal hosts, as well as KVM guests.
>
> For the example spinlock test, racing errors could not be reproduced in
> the default single-threaded TCG, even with a non-atomic lock. In KVM the
> expected behaviour of no errors with regular locks, and some errors with
> a non-atomic lock was observed. Next steps are to test multi-threaded TCG
> with this kind of payloads, extend the infrastructure to more complex and
> sensitive test cases, as well as support different architectures.
Can we reuse the kvm-unit-tests infrastructure? As you said, the
testcases should run just as well with TCG and KVM.
Paolo