Modifying htif.ld not affecting binaries

[Tim Liu]

Hi there!

We're trying to boot from our own memory, and have changed the new linker script - htif.ld. I've changed the line:

. = 0x80000000

to

. = 0x3000000

so that our instructions start at 0x3000000. However, when I compile my hello_world.c file to a .dump, the addresses still all start at 0x80000000. Is there another file or linker script that specifies where the code is placed in memory?

Thanks!

Tim

[Tim Liu]

Hi there,

I tried this again in Chipyard 1-2 and am running into the same problem. Any updates on where else the address needs to be changed?

Thanks!

Tim

[Colin Schmidt]

How are you doing the compilation?

Are you sure you are passing the linker script to the compiler?

--
You received this message because you are subscribed to the Google Groups "Chipyard" group.
To unsubscribe from this group and stop receiving emails from it, send an email to chipyard+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/chipyard/6cde4f71-16f6-4b2a-9cdf-706a2a10a88f%40googlegroups.com.

[Tim Liu]

Hi Colin, 

Thanks for the quick reply! And no, I am not sure I am passing the linker script to the compiler :'(

My compilation is just:

make hello_world.c

or:

make dumps hello_world.c

which is how I can see the addresses are wrong. Does something else have to be passed for htif.ld to be used?

Thanks!

Tim

On Friday, March 20, 2020 at 8:39:12 AM UTC-7, Colin Schmidt wrote:

How are you doing the compilation?

Are you sure you are passing the linker script to the compiler?

On Thu, Mar 19, 2020 at 9:00 PM Tim Liu <timl...@gmail.com> wrote:

Hi there,

I tried this again in Chipyard 1-2 and am running into the same problem. Any updates on where else the address needs to be changed?

Thanks!

Tim

On Friday, March 13, 2020 at 2:18:25 PM UTC-7, Tim Liu wrote:

Hi there!

We're trying to boot from our own memory, and have changed the new linker script - htif.ld. I've changed the line:

. = 0x80000000

to

. = 0x3000000

so that our instructions start at 0x3000000. However, when I compile my hello_world.c file to a .dump, the addresses still all start at 0x80000000. Is there another file or linker script that specifies where the code is placed in memory?

Thanks!

Tim

--
You received this message because you are subscribed to the Google Groups "Chipyard" group.

To unsubscribe from this group and stop receiving emails from it, send an email to chip...@googlegroups.com.

[Tim Liu]

Okay, I took a closer look at libgloss.mk. I've modified libgloss.mk to force a local build and set the target to riscv64-unknown-elf:

BUILD_LIBGLOSS=1

TARGET=riscv64-unknown-elf

and I'm seeing "Using local build" printed out and I can see that libgloss is:

libgloss/libgloss_htif.a ../toolchains/libgloss/util/htif_nano.specs htif.ld

So htif.ld is included in there. But running this make commands leaves the code at the original address. I've also tried adding 

-T htif.ld

to the command on the second line and the result is the same. Is there something else I'm missing to move the start address?

Thanks!

Tim

[Baltazar Ortiz]

Hi Tim,

Were you able to figure this out? I'm running into the same issue - opened an issue in the libgloss-htif repo in case that is a quicker turnaround as well.

[Albert Ou]

I had a fix on a separate branch, now pushed to master.

This changes the spec files to treat htif.ld as the default linker
script, which lets -T replace htif.ld entirely. Previously, since the
spec files used the -T option themselves, subsequent instances of -T
specified by the user merely appended to the existing script.

https://github.com/ucb-bar/libgloss-htif/commit/496a0b66b644241f1062e7c4ade87c55fa2fd286

> >>>> <https://groups.google.com/d/msgid/chipyard/6cde4f71-16f6-4b2a-9cdf-706a2a10a88f%40googlegroups.com?utm_medium=email&utm_source=footer>
> >>>> .

> >>>>
> >>>
>
> --
> You received this message because you are subscribed to the Google Groups "Chipyard" group.

> To unsubscribe from this group and stop receiving emails from it, send an email to chipyard+u...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/chipyard/f4284cdf-e3ad-48b6-a69f-e3ed0cd21396%40googlegroups.com.

--
Albert Ou

[Albert Ou]

Also, using a local version of htif.ld does not work if libgloss-htif
has already been installed to the toolchain location. Note that
directories specified with -L are added after the default linker search
path. Thus, ${RISCV}/riscv64-unknown-elf/lib/htif.ld takes precedence
over ./htif.ld in the current working directory.

--
Albert Ou

[Baltazar Ortiz]

Perfect, thank you for the fix and the clarification!