This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: [PATCH v3] sim: cfi: new flash device simulation
- From: "Frank Ch. Eigler" <fche at redhat dot com>
- To: Mike Frysinger <vapier at gentoo dot org>
- Cc: gdb-patches at sourceware dot org, toolchain-devel at blackfin dot uclinux dot org
- Date: Fri, 25 Mar 2011 15:29:11 -0400
- Subject: Re: [PATCH v3] sim: cfi: new flash device simulation
- References: <AANLkTi=FZ9TkZpRkWS2QQeQNBnEOCo-phavnaRx-79f1@mail.gmail.com>
Hi -
On Fri, Mar 25, 2011 at 03:19:28PM -0400, Mike Frysinger wrote:
> [...]
> > Woudln't you want to write(2) out the contents of the flash backing
> > store file, if !HAVE_MMAP?
>
> perhaps, but that'd require quite a bit more logic that just a single
> call to write() here. frankly, i dont care about the systems that
> dont support mmap as required by POSIX. the ifdef's are only to keep
> from breaking their builds. [...]
I see. If you don't want to implement write() or fwrite() in a
symmetric way to the read path, you could make sure that a !HAVE_MMAP
machine would reject the configuration of a cfi model in "rw"
file-backed mode.
- FChE