Mail Archives: djgpp-workers/2002/09/24/06:21:13
> > I didn't try this yesterday, but I did once I read the email and the
rm.exe
> > from simtel's fil41b.zip works fine with the example. I now expect that
the
> > problem is in a 2.04 CVS LIBC change or series of changes.
> [snip]
>
> I guess we already know that 2.03 and CVS are just as broken as each
other,
> but I got error messages about the inodes yesterday, when using rm.exe
under
> Windows NT 4 (without LFN TSR) under VMware 3.2.
>
> So here's a proposed fix:
>
> * We add a hack (#ifdef/#endif) to Fileutils 4.1, to not check for the
inode
> number changing, for DJGPP 2.03. From your earlier posts, Andrew, it
sounds
> like this may not work.
I tried to #ifdef/#endif out the assert and the check and I seemed to get
into some sort of recursive loop in rm.exe as it never stopped until I
CTRL-C'd it. I may not have #ifdef/#endif the right bit.
> * We fix DJGPP CVS to calculate the inode the same way, so that the check
> doesn't fail.
It looks like lstat() returns the correct inode (same), but I need to spend
some more time debugging remove.c
> I didn't have any time to look at this issue properly at the weekend.
> Hopefully I will have some time this weekend.
That makes two of us.
Andrew
- Raw text -