Messages in this thread | | | From | Grant Grundler <> | Subject | [PATCH 0/5] net: atlantic: more fuzzing fixes | Date | Mon, 18 Apr 2022 16:17:41 -0700 |
| |
The Chrome OS fuzzing team posted a "Fuzzing" report for atlantic driver in Q4 2021 using Chrome OS v5.4 kernel and "Cable Matters Thunderbolt 3 to 10 Gb Ethernet" (b0 version): https://docs.google.com/document/d/e/2PACX-1vT4oCGNhhy_AuUqpu6NGnW0N9HF_jxf2kS7raOpOlNRqJNiTHAtjiHRthXYSeXIRTgfeVvsEt0qK9qK/pub
It essentially describes four problems: 1) validate rxd_wb->next_desc_ptr before populating buff->next 2) "frag[0] not initialized" case in aq_ring_rx_clean() 3) limit iterations handling fragments in aq_ring_rx_clean() 4) validate hw_head_ in hw_atl_b0_hw_ring_tx_head_update()
I've added one "clean up" contribution: "net: atlantic: reduce scope of is_rsc_complete"
I tested the "original" patches using chromeos-v5.4 kernel branch: https://chromium-review.googlesource.com/q/hashtag:pcinet-atlantic-2022q1+(status:open%20OR%20status:merged)
The fuzzing team will retest using the chromeos-v5.4 patches and the b0 HW.
I've forward ported those patches to 5.18-rc2 and compiled them but am currently unable to test them on 5.18-rc2 kernel (logistics problems).
I'm confident in all but the last patch: "net: atlantic: verify hw_head_ is reasonable"
Please verify I'm not confusing how ring->sw_head and ring->sw_tail are used in hw_atl_b0_hw_ring_tx_head_update().
Credit largely goes to Chrome OS Fuzzing team members: Aashay Shringarpure, Yi Chou, Shervin Oloumi
cheers, grant
| |