linux-kernel - A bug or a feature of CONFIG_HZ on a low clock frequency FPGA

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: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [day] [month] [year] [list]

Message-ID: <CAO0Zsh+OT_JNZC8OV_cG7LKn2y2ZYM2hm=SAapW5FEub-qa_BA@mail.gmail.com>
Date:   Fri, 18 Nov 2022 17:31:38 +0800
From:   赵夏 <zhaoxiahust@...il.com>
To:     linux-kernel@...r.kernel.org
Subject: A bug or a feature of CONFIG_HZ on a low clock frequency FPGA

Hi guys,

  We are running the linux kernel on a FPGA board with 20MHz clock
frequency. If we configure the CONFIG_HZ as 12, the kernel executes
perfectly. However, after configuring the CONFIG_HZ to 250 as a test,
the kernel boot stucks after printing "rcu: Hierarchical SRCU
implementation".

  We added one printk sentence in the timer_interrupt function to
check the next instruction pc stored in regs. It keeps printing the pc
of the replay_soft_interrupts function.

  We guess this is because the timer interrupt happens again when the
kernel is still processing the previous timer interrupt. Thus, the
kernel jmps into timer_interrupt again and again.

  If we are right, are there any guidelines to configure the CONFIG_HZ
based on the hardware platform clock frequency? If we are wrong, how
do we explain the stucking case that we met?

Thanks
Xia