From: Alexander Bluhm Subject: Re: igc(4): fix recv. jumbo frames To: David Gwynne Cc: Jan Klemkow , tech@openbsd.org, Theo de Raadt Date: Fri, 9 Aug 2024 11:52:00 +0200 On Fri, Aug 09, 2024 at 11:55:35AM +1000, David Gwynne wrote: > On Thu, Aug 08, 2024 at 12:51:18PM +0200, Jan Klemkow wrote: > > Index: dev/pci/if_igc.c > > =================================================================== > > RCS file: /data/mirror/openbsd/cvs/src/sys/dev/pci/if_igc.c,v > > diff -u -p -r1.25 if_igc.c > > --- dev/pci/if_igc.c 24 May 2024 06:02:53 -0000 1.25 > > +++ dev/pci/if_igc.c 6 Aug 2024 08:36:31 -0000 > > @@ -881,7 +881,7 @@ igc_init(void *arg) > > } > > igc_initialize_transmit_unit(sc); > > > > - sc->rx_mbuf_sz = MCLBYTES + ETHER_ALIGN; > > + sc->rx_mbuf_sz = MCLBYTES; > > i think this is a mistake. You are right, it immediately crashes on sparc64. Sorry for not testing before. bluhm login: panic: trap type 0x34 (mem address not aligned): pc=16dc8c0 npc=16dc8c4 pstate=820006 Stopped at db_enter+0x8: nop TID PID UID PRFLAGS PFLAGS CPU COMMAND *276857 66236 0 0x14000 0x200 2 softnet1 trap(403be16f6c0, 34, 16dc8c0, 820006, 0, 6) at trap+0x334 Lslowtrap_reenter(4004e3e5100, 0, 20, 0, 403be2cdcc4, 403be2cde00) at Lslowtrap_reenter+0xf8 mtx_enter_try(4004aa9b048, 4004e3e5100, 1837b40, 1c7d290, 12deb00, 0) at mtx_enter_try+0x8c ip_input_if(403be16fa78, 403be16fa84, 4, 0, 4004aa9b048, 12) at ip_input_if+0x78 ipv4_input(4004aa9b048, 19b4148, 4004aa9b050, 1, 62, 0) at ipv4_input+0x3c ether_input(4004aa9b048, 4004e3e5100, 17c3578, 0, 0, 17c3578) at ether_input+0x274 if_input_process(4004aa9b048, 403be16fcf0, 20, 17c3578, 0, 6) at if_input_process+0x44 ifiq_process(4004aa97c00, 403be16fdd8, 1c86830, 1c7d290, 6, 6) at ifiq_process+0x74 taskq_thread(4004a97a080, 4004a9187e0, 403be16ba78, 3, 0, 6) at taskq_thread+0x7c proc_trampoline(0, 0, 0, 0, 0, 0) at proc_trampoline+0x10 https://www.openbsd.org/ddb.html describes the minimum info required in bug reports. Insufficient info makes it difficult to find and fix bugs. ddb{2}>