[PATCH v5 3/6] firmware: Explicitly pass -pie to the linker, not just the driver

[Jessica Clarke]

On 11 Jul 2021, at 14:05, Bin Meng <bmeng.cn at gmail.com> wrote:
> 
> On Sun, Jul 11, 2021 at 10:29 AM Jessica Clarke <jrtc27 at jrtc27.com> wrote:
>> 
>> When using Clang with a bare-metal triple, -pie does not get passed to
>> the linker as it's not normally a thing that makes sense, unlike GCC
>> which will unconditionally forward it on and potentially result in a
> 
> It seems GCC driver does not unconditionally forward it to the linker, see:
> 
> $ riscv64-unknown-elf-gcc -fPIE -nostdlib -x c /dev/null -o /dev/null
> 2>&1 && echo y
> /opt/riscv-unknown-elf/lib/gcc/riscv64-unknown-elf/10.1.0/../../../../riscv64-unknown-elf/bin/ld:
> warning: cannot find entry symbol _start; defaulting to
> 0000000000010040
> y
> 
> But
> 
> $ riscv64-unknown-elf-gcc -fPIE -nostdlib -Wl,-pie -x c /dev/null -o
> /dev/null 2>&1 && echo y
> /opt/riscv-unknown-elf/lib/gcc/riscv64-unknown-elf/10.1.0/../../../../riscv64-unknown-elf/bin/ld:
> -pie not supported
> collect2: error: ld returned 1 exit status
> 
> So regarding bare-metal triple, GCC driver and  clang driver behave
> the same way. It's just GNU bare-metal ld does not support PIE, while
> LLD supports.
> 
> So we need to revise the above commit message.

You didn’t pass -pie, only -fPIE.

Jess

>> linker error. However, LLD does support it, and manually forwarding it
>> on works as desired, so do so to fully support FW_PIC with Clang and
>> LLD.
>> 
>> Signed-off-by: Jessica Clarke <jrtc27 at jrtc27.com>
>> ---
>> firmware/objects.mk | 2 +-
>> 1 file changed, 1 insertion(+), 1 deletion(-)
>> 
> 
> Regards,
> Bin