collect errors out with "Error: invalid digit found in string"

[pspacek]

Summary

tools/retis_in_container.sh collect errors out with "Error: invalid digit found in string"

Environment

• kernel 6.4.12-arch1-1
• Docker container: quay.io/retis/retis 4333f006997c
• docker on the host: 1:24.0.5-1
• host is Arch Linux on x86_64 updated as of 2023-08-28

Steps to reproduce

• tools/retis_in_container.sh collect

Output

Error: invalid digit found in string

I've tried to add --log-level debug but it does not extend the output.

I'm at loss what else to provide, but I'm happy to add more info if you tell me what's needed.

[pspacek]

I have big trouble debugging this using GDB or LLDB because GDB crashes on me constantly, and LLDB does not show me the right level of detail (but I admit I'm using it for the first time ...).

Anyway, I sense this might have something to do with Inspector constructor.

Content of /etc/os-release on my system:

NAME="Arch Linux"
PRETTY_NAME="Arch Linux"
ID=arch
BUILD_ID=rolling
ANSI_COLOR="38;2;23;147;209"
HOME_URL="https://archlinux.org/"
DOCUMENTATION_URL="https://wiki.archlinux.org/"
SUPPORT_URL="https://bbs.archlinux.org/"
BUG_REPORT_URL="https://bugs.archlinux.org/"
PRIVACY_POLICY_URL="https://terms.archlinux.org/docs/privacy-policy/"
LOGO=archlinux-logo

parse_os_info() returns: Ok Some("arch"), None

[pspacek]

Poking at it bit more, it seems the last line it executes is

let version = KernelVersion::new()?;

GDB is crashing on me so I'm not able tell what exactly is going on here.

Anyway,

$ uname -a
Linux p 6.4.12-arch1-1 #1 SMP PREEMPT_DYNAMIC Thu, 24 Aug 2023 00:38:14 +0000 x86_64 GNU/Linux

[pspacek]

On line 51 Ok(UtsName(ret.assume_init())) the ret value is:

(core::mem::maybe_uninit::MaybeUninit<libc::unix::linux_like::utsname>) $13 = {
  value = {
    value = {
      sysname = "Linux"
      nodename = "p"
      release = "6.4.12-arch1-1"
      version = "#1 SMP PREEMPT_DYNAMIC Thu, 24 Aug 2023 00:38:14 +0000"
      machine = "x86_64"
      domainname = "(none)"
    }
  }
}

and indeed it seems that the kernel version is what's causing the explosion.

[amorenoz]

@pspacek Thanks for reporting and for the initial analysis. Indeed, it seems the kernel version convention used in Arch is not being parsed properly.

[amorenoz]

Still trying to figure out if this is the right solution, but in the time being, could you please check if this fixes your issue? https://github.com/retis-org/retis/pull/242