On 10/4/2010 4:14 PM, Wolfgang Denk wrote: > Dear Vipin Kumar, > > In message <4ca9acaa.2020...@st.com> you wrote: >> >>>> This writel results in writing byte by byte on the address pointed to by >>>> status_reg. >>>> This problem is visible with both gcc version 4.4.1 as well as 4.5.0 >>> >>> I bet this is on some ARM system? >> >> Yes, it is on an ARM system (CortexA9). But I still feel that since I am >> creating >> a new u32 * status_reg, the code should not use any intelligence and use the >> pointer >> only to produce an str instruction in the form >> str r0, [r1] >> >> But it retains the packed property of the structure even with a new u32 >> ponter >> typecasted to u32 * >> u32 * status_reg = (u32 *)xyz->x; >> >> A writel to status_reg results in byte by byte writing > > I agree with you. I always considered such behaviour of the ARM C > compiler a bug, and still do. However, people with better knowledge > of the ARm architecture than me might be able to explain why the > responsible PTB consider this to be a good and necessary "feature" of > th compiler. > >>> Hm... Why do these structs have any "__attribute__ ((packed))" at all? >> >> Even I could not understand that very well > > Eventually alignment of these structs cannot be guaranteed? >
In my opinion it can be guaranteed. btw, I am talking about ehci_hcor structure in include/usb/host/ehci.h The only reason I am confused is that a lot many platforms would have faced a similar problem (or is it only me). Please confirm if I should remove the packed attribute and send a patch This also raises one doubt. Since u-boot code now contains structures to access device registers, using packed attribute with these structures can be lethal Regards Vipin _______________________________________________ U-Boot mailing list U-Boot@lists.denx.de http://lists.denx.de/mailman/listinfo/u-boot
