Date: Fri, 06 Jun 2008 08:40:10 +0100 From: "Bruce M. Simpson" <bms@FreeBSD.org> To: =?ISO-8859-1?Q?Marc_L=F6rner?= <marc.loerner@hob.de> Cc: freebsd-net@freebsd.org, Rui Paulo <rpaulo@FreeBSD.org> Subject: Re: Probable Bug in tcp.h Message-ID: <4848E9DA.8090802@FreeBSD.org> In-Reply-To: <200806060930.28527.marc.loerner@hob.de> References: <200806051712.47048.marc.loerner@hob.de> <20080605155646.GC6864@epsilon.local> <200806060930.28527.marc.loerner@hob.de>
next in thread | previous in thread | raw e-mail | index | archive | help
Marc Lörner wrote: > th_x2 and th_off are created as a bitfield. But C-Standard says that bitfields > are accessed as integers => 4-bytes > > On itanium integers are read with ld4-command but the address of th_x2/th_off > may not be aligned to 4-bytes => we get an unaligned reference fault. > > If we'd change to 1 byte-accesses => I won't get any misaligned faults > anymore. > It's worth noting that Linux implements its version of tcphdr using a 32-bit-wide bitfield and the TCP header flags live there as bits instead of as integer quantities. I think it should be OK to change the u_int to a uint8_t as NetBSD has. The problem with bitfields in "signed char" is that they can become unintentionally sign extended on a read, and for many years compilers only supported "char", not "unsigned char". Does anyone see a reason why we should not make this change?
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4848E9DA.8090802>