Patch / RFC: add "flashcp" applet

[Stefan Seyfried]

On Sat, 14 Nov 2009 23:03:02 +0100
Bernhard Reutner-Fischer <rep.dot.nop at gmail.com> wrote:

> On Sat, Nov 14, 2009 at 08:00:44PM +0100, Stefan Seyfried wrote:
> >Hi,
> >
> >the attached patch implements a "flashcp" similar the one form
> >mtd-utils.
> >
> >I have read the style-guide and I think that I did everything right.
> >If so, please "git am", if not, please comment ;)
> 
>   CC      miscutils/flashcp.o
> miscutils/flashcp.c: In function ‘flashcp_main’:
> miscutils/flashcp.c:121: warning: format ‘%.8x’ expects type ‘unsigned
> int’, but argument 2 has type ‘size_t’
> miscutils/flashcp.c:121: warning: format ‘%.8x’ expects type ‘unsigned
> int’, but argument 3 has type ‘size_t’
> miscutils/flashcp.c:121: warning: format ‘%d’ expects type ‘int’, but
> argument 5 has type ‘long unsigned int’
> miscutils/flashcp.c:128: warning: format ‘%.8x’ expects type ‘unsigned
> int’, but argument 2 has type ‘size_t’
> miscutils/flashcp.c:128: warning: format ‘%.8x’ expects type ‘unsigned
> int’, but argument 3 has type ‘size_t’

hm, did not happen to me on x86 and ppc405, but maybe I have the wrong
warning options set.

> $ ./scripts/bloat-o-meter _bb_un.orig busybox_unstripped
> function                                             old     new   delta
> flashcp_main                                           -     835    +835
> packed_usage                                       26028   26128    +100
> progress                                               -      67     +67
> applet_names                                        2102    2110      +8
> applet_main                                         2496    2504      +8
> applet_nameofs                                       624     626      +2
> applet_install_loc                                   156     157      +1
> ------------------------------------------------------------------------------
> (add/remove: 2/0 grow/shrink: 5/0 up/down: 1021/0)           Total: 1021
> bytes
> 
> $ size miscutils/flashcp.o 
>    text	   data	    bss	    dec	    hex	filename
>    1344	      0	      0	   1344	    540	miscutils/flashcp.o
> 
> 1.3k
> don't get me wrong, but this applet seems to:
> -) loop through mtd, eraseing 1k
> -) loop through file, bb_copyfd_exact_size()

but bb_copyfd_exact_size does not do progress indication. And I know
nobody who uses flashcp without "-v" parameter ;) (usually, it is a
quite time consuming operation, in the order of tens of seconds to a
few minutes, so progress indication is really useful)

>    and verify bytes re-read do memcmp==0 bytes written
> 
> Sounds more like 600b-800b than more than 1k, no?

To be honest, I have no real feeling for how big an applet should be,
but I trust you on your numbers, and thus I will try to squeeze a few
more bytes out of it ;)

> I encourage you to license your new applet under GPLv2 *or later* so we
> don't have to rewrite it later on (eventually).

That's no problem. Even though I am pretty reluctant to trust the FSF
on future decisions, this piece of Code is so trivial that it might as
well be licensed under the WTFPL v2 ;-)

> Anyway. Attached patch would kill off 100b, but i think rewriting it
> from scratch would be easier.

Just one note:

+	filename = *argv++;
+	devicename = *argv;

I had something like that before (instead of argv[0]), but it resulted
in significantly bigger code.
I did not understand why, but it was definitely true.

-- 
Stefan Seyfried

"Any ideas, John?"
"Well, surrounding them's out."