This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: End markers for functions.
- To: Geoff Keating <geoffk at ozemail dot com dot au>
- Subject: Re: End markers for functions.
- From: Andrew Morton <akpm at zip dot com dot au>
- Date: Wed, 17 Nov 1999 11:25:09 +0000
- CC: gcc-patches at gcc dot gnu dot org
- References: <199911171024.VAA01137@gluttony.geoffk.wattle.id.au>
Geoff Keating wrote:
>
> +/* -ffunction-end-marker=ASM causes ASM to be emitted after the
> + function epilogue. This can be used to help runtime systems to
> + construct a stack-trace for architectures that don't maintain a
> + frame pointer chain. At stack-trace time, the runtime system scans
> + backward from the PC to find the end of the previous function (and
> + thus the beginning of this one), and from there find and decode the
> + prologue. The end of the function is marked because control flow
> + can, on some systems, fall into the start of the function at the
> + beginning of the file. */
> +
Interested.
What advantage does this have over searching the symtab for the
highest-address function symbol which is less than the PC?
Is this feature driven by a GDB requirement or by some other application
which you had in mind?
Thanks.