Ming Zhang wrote: > On Tue, 2006-12-19 at 18:00 +0200, Momchil Velikov wrote: >> Ming Zhang wrote: >>> On Tue, 2006-12-19 at 12:56 +0200, Momchil Velikov wrote: >>>> Arjan van de Ven wrote: >>>>> On Mon, 2006-12-18 at 17:12 -0500, Ming Zhang wrote: >>>>>> See this code piece >>>>>> >>>>>> http://lxr.linux.no/source/drivers/scsi/libata-scsi.c?v=2.6.18#L1049 >>>>>> >>>>>> 1056 lba |= ((u64)scsicmd[2]) << 24; >>>>>> 1057 lba |= ((u64)scsicmd[3]) << 16; >>>>>> 1058 lba |= ((u64)scsicmd[4]) << 8; >>>>>> 1059 lba |= ((u64)scsicmd[5]); >>>>>> >>>>>> it can also be written as >>>>>> >>>>>> lba = be32_to_cpu(*(u32 *)(&scsicmd[2]) >>>> ISO/IEC 9899:1999(E) "6.5 Expressions" [#7] >>> read but still can not understand why u list this here. >> Because this code >> >> ba = be32_to_cpu(*(u32 *)(&scsicmd[2]) >> >> violates this requirement >> >> ISO/IEC 9899:1999(E) "6.5 Expressions" [#7] >> >> unless you have checked in all callers of ``scsi_10_lba_len'' (and >> transitively in their callers, if necessary) that the first parameter in >> fact does refer to an object of type ``u32'', in which case the above >> code becomes simply non-robust instead of outright non-conforming. >> > > scsicmd is u8*, so scsicmd + 2 is also u8*, why can not cast it to u32*? > u* is a unsigned char so a character type" right? Who said you cannot cast it? The referred to text in the standard speaks about *access* to object's stored value, not casts. ~velco -- Kernelnewbies: Help each other learn about the Linux kernel. Archive: http://mail.nl.linux.org/kernelnewbies/ FAQ: http://kernelnewbies.org/faq/
