git-rebase (1.5.0.6) errors

Hi,

ijjarvin@kivilampi-30:~/work/src/submit$ git-rebase net-2.6.22-origin First, rewinding head to replay your work on top of it... HEAD is now at d739437... [IPV4]: Correct rp_filter help text. fatal: cannot convert from utf-8 to utf-8

ijjarvin@kivilampi-30:~/work/src/submit$ git-log -n 1 | cat
commit d739437207064cdcea8f9c81442284106cbcb67f
Author: Dave Jones <davej@redhat.com>
Date:   Thu May 17 15:02:21 2007 -0700

    [IPV4]: Correct rp_filter help text.
    
    As mentioned in http://bugzilla.kernel.org/show_bug.cgi?id=5015
    The helptext implies that this is on by default.
    This may be true on some distros (Fedora/RHEL have it enabled
    in /etc/sysctl.conf), but the kernel defaults to it off.
    
    Signed-off-by: Dave Jones <davej@redhat.com>
    Signed-off-by: David S. Miller <davem@davemloft.net>
ijjarvin@kivilampi-30:~/work/src/submit$ git-rebase --abort
HEAD is now at e413863... [TCP] FRTO: Add missing ECN CWR sending to one of the responses
ijjarvin@kivilampi-30:~/work/src/submit$ git --version
git version 1.5.0.6

I find at least two problems in it...
  - Non-sense error message: "convert from utf-8 to utf-8" ?!?
  - It fails to rebase the e413863 changeset

2007/5/18, Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>:

Now I can't test or try to reproduce your error, but looking at the source code (only with gitweb) I have found another bad use of size_t instead of ssize_t in the reencode_string(..) that take part at the conversion process.

Using size_t in the next portion of code the check "count == -1" is never true.

while (1) {
		size_t cnt = iconv(conv, &cp, &insz, &outpos, &outsz);

		if (cnt == -1) {
			size_t sofar;
			if (errno != E2BIG) {
				free(out);
				iconv_close(conv);
				return NULL;
			}

Please someone could fixes this bad use of size_t?..

Sorry, but now I can't install/use git because I'm on a customer workstation..

"Paolo Teti" <paolo.teti@gmail.com> writes:

Only if size_t is a larger type than int (could be on x86-64 and alpha architectures). Other than that, this comparison would work. Which does not mean that this does not warrant fixing, but it is not necessarily the cause of this problem.

2007/5/18, David Kastrup <dak@gnu.org>:

1. Yes this is not necessarily the cause.. I agree.

2. size_t as I know is unsigned and ssize_t is the signed version of
size_t.. so to
    compare with -1 we want ssize_t.

David Kastrup <dak@gnu.org> wrote:

...sizeof(size_t) == sizeof(int) should hold...

Anyway, if this has any relevance: I'm using non-utf system, and (as you see) my surname has ä... The system was recently upgraded to git 1.5+ which started to complain also about a missing i18n.commitencoding, figured out that when I set it to utf8 (empty => defaults to it) and have signed-off line (with native non-utf ä), I get that error...

...and please, do not drop me from cc since I'm not subscribed...

2007/5/18, Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>:

what I have reported is not necessarily the cause of your error. Yes

BUT size_t is UNSIGNED and ssize_t is SIGNED.

The problem is that We store a signed return value into a unsigned variable. So the check against -1 has no sense.

and please note that also sizeof(int) == sizeof(unsigned int), but -1 is a signed int !!

I repeat: this is not necessarily the cause of your error.

On Fri, May 18, 2007 at 17:02:56 +0300, Ilpo Järvinen wrote:

Really?

$ cat test.c
#include <stdio.h>
int main(void)
{
    printf("sizeof(int) = %i\n", sizeof(int));
    printf("sizeof(long) = %i\n", sizeof(long));
    printf("sizeof(size_t) = %i\n", sizeof(size_t));
    return 0;
}
$ gcc -otest test.c
$ ./test
sizeof(int) = 4
sizeof(long) = 8
sizeof(size_t) = 8

Hm, it does not seem that sizeof(size_t) == sizeof(int).

$ uname -m x86_64

Yes, this is a 64-bit system.

Anyway, comparing it with -1 is ALWAYS OK in spite of this!

$ cat test2.c
#include <stdio.h>
int main(void)
{
    size_t x = 0;
    --x;
    printf("x = 0x%lx\n", x);
    printf("(x == -1) = %i\n", (x == -1));
    return 0;
}
$ gcc -otest2 test2.c
$ ./test2
x = 0xffffffffffffffff
(x == -1) = 1

So at least with gcc that comparison is OK anyway. There has to be something else that causes that problem.

On Sun, 20 May 2007, Jan Hudec wrote:

...On 64-bit perhaps but mine isn't one of them... I'm sorry if my wording was misleading, I meant to say that on my system the equality of sizeof()s should hold... :-)

ijjarvin@kivilampi-30:~/src/testsize$ ./main sizeof(int) = 4 sizeof(long) = 4 sizeof(size_t) = 4

$ uname -m i686

2007/5/18, David Kastrup <dak@gnu.org>:

Stupid me.. OK is == -1 and not < 0..

Probably today  I'm drunk without drink ..

Enjoy