Date: Wed, 7 Sep 2016 07:22:59 -0600 From: Warner Losh <imp@bsdimp.com> To: Erik Moe <e.moe@rcn.com> Cc: freebsd-arm <freebsd-arm@freebsd.org> Subject: Re: Unaligned access in ubldr.bin Message-ID: <CANCZdfqBH5w2SYaiuvdgaZ0DZTBLjSdTAcye4BL%2B3NhaY=f6sA@mail.gmail.com> In-Reply-To: <E04431D8-F38C-4B9F-BFE1-1207265DED8B@rcn.com> References: <E04431D8-F38C-4B9F-BFE1-1207265DED8B@rcn.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, Sep 7, 2016 at 12:46 AM, Erik Moe <e.moe@rcn.com> wrote:
> Hello,
>
> I=E2=80=99ve been trying to bring up FreeBSD on the USB Armory based on t=
he i.MX53. I=E2=80=99ve made myself a u-boot patch that will load and star=
t ubldr.bin but I=E2=80=99m running into a issue with what I think is an un=
aligned access issue:
which version of u-boot are you using?
> ## Starting application at 0x70800000 ...
> data abort
> pc : [<70824bac>] lr : [<7081846c>]
> sp : 8f550c98 ip : 70835a00 fp : 8f550cb0
> r10: 00000002 r9 : 70838d58 r8 : 70833cd9
> r7 : 707fff08 r6 : 000054f0 r5 : 70833cd9 r4 : 00000000
> r3 : 70828fc4 r2 : 70833cd9 r1 : 00000001 r0 : 7083705c
> Flags: Nzcv IRQs off FIQs off Mode SVC_32
> Resetting CPU ...
>
> I=E2=80=99ve hand dissambled the code at pc=3D0x70824bac:
>
> 0x70824ba8: 0xe59f21d0 ldr r2, [pc, #464]
> 0x70824bac: 0xe5825000 str r5, [r2]
> 0x70824bb0: 0xe5d53000 ldrb r3, [r5]
> 0x70824bb4: 0xe353002d cmp r3, #45
> 0x70824bb8: 0x1a00000b bne #+44
>
> The offending instructions is "str r5, [r2]=E2=80=9D where r2 =3D 0x70833=
cd9, which sort of makes sense since it isn=E2=80=99t aligned on a 4 byte b=
oundary. I=E2=80=99m new to arm ARM, so I=E2=80=99m not really sure. My q=
uestions are these:
>
> 1.) Doesn=E2=80=99t ARMv6 and higher architecture allow unaligned access?=
Is there something that u-boot needs be doing in initialization to allow =
unaligned access?
It does, but usually that's configured later in boot.
> 2.) Does ubldr make the assumption that unaligned access is allowed and m=
aybe shouldn=E2=80=99t? I would think not since ubldr has been around for =
a while and works on numerous ARM processors.
It may be that the compiler is generating bad code in this case? You
might check to see how we're building it and to see if that's causing
problems.
I've been trying to bring up v2016.07 on some boards and that fails
where v2016.05 worked. I get data alignment issues, but all I'm
changing is u-boot.img.
Warner
> Any thoughts at which direction to investigate would be appreciated.
>
> Thanks,
> Erik
>
> FYI, ubldr is failing in =E2=80=9Cgetopt=E2=80=9D. I=E2=80=99ve used obj=
dump to disassemble ubldr.bin. You=E2=80=99ll have to subtract the offset =
0xf8 and add 0x7080000 to the addresses bellow to get them to align with th=
e addresses above.
>
>
> getopt(int nargc, char * const *nargv, const char *ostr)
> {
> 24c4c: e92d4df0 push {r4, r5, r6, r7, r8, sl, fp, lr}
> 24c50: e28db018 add fp, sp, #24
> 24c54: e1a07000 mov r7, r0
> static char *place =3D EMSG; /* option letter proces=
sing */
> char *oli; /* option letter list ind=
ex */
>
> if (optreset || !*place) { /* update scanning pointe=
r */
> 24c58: e59f0214 ldr r0, [pc, #532] ; 24e74 <getopt+0=
x228>
> 24c5c: e1a06001 mov r6, r1
> 24c60: e1a08002 mov r8, r2
> 24c64: e5901000 ldr r1, [r0]
> 24c68: e3510000 cmp r1, #0
> 24c6c: 1a000004 bne 24c84 <getopt+0x38>
> 24c70: e59f1200 ldr r1, [pc, #512] ; 24e78 <getopt+0=
x22c>
> 24c74: e5913000 ldr r3, [r1]
> 24c78: e5d34000 ldrb r4, [r3]
> 24c7c: e3540000 cmp r4, #0
> 24c80: 1a00001d bne 24cfc <getopt+0xb0>
> 24c84: e3a01000 mov r1, #0
> optreset =3D 0;
> 24c88: e5801000 str r1, [r0]
> if (optind >=3D nargc || *(place =3D nargv[optind]) !=3D =
'-') {
> place =3D EMSG;
> return (-1);
> }
> if (place[1] && *++place =3D=3D '-') { /* found "--"=
*/
> ++optind;
> 24c8c: e59f01e8 ldr r0, [pc, #488] ; 24e7c <getopt+0=
x230>
> 24c90: e5901000 ldr r1, [r0]
> static char *place =3D EMSG; /* option letter proces=
sing */
> char *oli; /* option letter list ind=
ex */
>
> if (optreset || !*place) { /* update scanning pointe=
r */
> optreset =3D 0;
> if (optind >=3D nargc || *(place =3D nargv[optind]) !=3D =
'-') {
> 24c94: e1510007 cmp r1, r7
> 24c98: aa000011 bge 24ce4 <getopt+0x98>
> 24c9c: e7965101 ldr r5, [r6, r1, lsl #2]
> 24ca0: e59f21d0 ldr r2, [pc, #464] ; 24e78 <getopt+0=
x22c>
> 24ca4: e5825000 str r5, [r2]
> 24ca8: e5d53000 ldrb r3, [r5]
> 24cac: e353002d cmp r3, #45 ; 0x2d
>
>
> _______________________________________________
> freebsd-arm@freebsd.org mailing list
> https://lists.freebsd.org/mailman/listinfo/freebsd-arm
> To unsubscribe, send any mail to "freebsd-arm-unsubscribe@freebsd.org"
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CANCZdfqBH5w2SYaiuvdgaZ0DZTBLjSdTAcye4BL%2B3NhaY=f6sA>