When did CVE-2025-38280 actually become a problem?
Hi all,
I sent basically the same question to [email protected] but they are out of ideas. They assign the affected version numbers based on the "Fixes" information initially. But I'm unsure if that one is actually correct here, see below.
The fix is this commit:
Quoted 6 lines
commit
86bc9c742426a16b52a10ef61f5b721aecca2344Author: KaFai Wan <[email protected]> Date: Mon May 26 21:33:58 2025 +0800bpf: Avoid
__bpf_prog_ret0_warnwhen jit fails
[…]
Quoted 1 line
Fixes fa9dd599b4da ("bpf: get rid of pure_initcall dependency to enable
jits")
And my questions were those:
=========
I was staring a while on CVE-2025-38280, especially since the message states:
Quoted 2 lines
When creating bpf program, '
fp->jit_requested' depends onbpf_jit_enable. This issue is triggered because ofCONFIG_BPF_JIT_ALWAYS_ONis not set …
But the commit that this was attributed to (5124abda3060e2eab506fb14a27acadee3c3e396) added the warning to the code, but the function is only reachable when CONFIG_BPF_JIT_ALWAYS_ON is set. This was the case until 6ebc5030e0c5a698f1dd9a6684cddf6ccaed64a0 moved it out of the define. So is this even an issue before 6.15 after all? Since the fix got backported I think it's more an issue to where the second commit got backported? So in my eyes the 5.10 kernel I'm currently staring at isn't affected at all.
==========
Can anyone comment on this? If there is a conclusion I can relay that to the CVE folks to update the version ranges afterwards.
Regards,
Eike
Signature · 13 lines
Rolf Eike Beer emlix GmbH Headquarters: Berliner Str. 12, 37073 Göttingen, Germany Phone +49 (0)551 30664-0, e-mail [email protected] District Court of Göttingen, Registry Number HR B 3160 Managing Directors: Heike Jordan, Dr. Uwe Kracke VAT ID No. DE 205 198 055 Office Berlin: Panoramastr. 1, 10178 Berlin, Germany Office Bonn: Bachstr. 6, 53115 Bonn, Germany http://www.emlix.com emlix - your embedded Linux partner
Quoted 37 lines · Rolf Eike Beer
Hi all,
I sent basically the same question to [email protected] but they are out of ideas. They assign the affected version numbers based on the "Fixes" information initially. But I'm unsure if that one is actually correct here, see below.
The fix is this commit:
Quoted 6 lines
commit
86bc9c742426a16b52a10ef61f5b721aecca2344Author: KaFai Wan <[email protected]> Date: Mon May 26 21:33:58 2025 +0800bpf: Avoid
__bpf_prog_ret0_warnwhen jit fails[…]
Quoted 1 line
Fixes fa9dd599b4da ("bpf: get rid of pure_initcall dependency to enablejits")
And my questions were those:
=========
I was staring a while on CVE-2025-38280, especially since the message states:Quoted 2 lines
When creating bpf program, '
fp->jit_requested' depends onbpf_jit_enable. This issue is triggered because ofCONFIG_BPF_JIT_ALWAYS_ONis not set …But the commit that this was attributed to (
5124abda3060e2eab506fb14a27acadee3c3e396) added the warning to the code, but the function is only reachable whenCONFIG_BPF_JIT_ALWAYS_ONis set. This was the case until6ebc5030e0c5a698f1dd9a6684cddf6ccaed64a0moved it out of the define. So is this even an issue before 6.15 after all? Since the fix got backported I think it's more an issue to where the second commit got backported? So in my eyes the 5.10 kernel I'm currently staring at isn't affected at all.
==========
Quoted 2 lines
Can anyone comment on this? If there is a conclusion I can relay that to the CVE folks to update the version ranges afterwards.
Agree with the analysis. As far as the kernel warning is concerned, I don't think it can happen before commit 6ebc5030e0c5.
Best,
Shung-Hsi
Quoted 17 lines
Regards,
Eike
Signature · 13 lines
Rolf Eike Beer emlix GmbH Headquarters: Berliner Str. 12, 37073 Göttingen, Germany Phone +49 (0)551 30664-0, e-mail [email protected] District Court of Göttingen, Registry Number HR B 3160 Managing Directors: Heike Jordan, Dr. Uwe Kracke VAT ID No. DE 205 198 055 Office Berlin: Panoramastr. 1, 10178 Berlin, Germany Office Bonn: Bachstr. 6, 53115 Bonn, Germany http://www.emlix.com emlix - your embedded Linux partner