>>>>> "Russell" == Russell King <- ARM Linux Admin <[EMAIL PROTECTED]>> writes:
Russell> Paul Koning writes:
>> There was a discussion on this list only a week or two ago about a
>> bug in the glibc password handling functions -- caused by an
>> invalid assumption about alignment of data elements in structures.
>> Solution was to mark the structure in question "packed". A scan
>> of the archives should turn it up soon enough...
Russell> Ok, this really starts to worry me...
Russell> Why would something that reads a textual file require it's
Russell> structure to be packed? I don't get this.
Russell> Ok, so adding the stuff to make the structure packed fixed
Russell> the problem, but I think someone ought to fix it the 'right
Russell> way' and find out why this structure needs to be packed when
Russell> it shouldn't need to be.
I don't remember the specifics, and I don't have the original note
anymore. I do have one message from that thread. The subject was
"Patches to glibc...."; Scott Bambrough was one of the others in the
discussion. Date I have is around April 14th.
>From vague recollection, I think the issue had to do with a data
structure being passed around in which the request for a lookup was
encoded. Yes of course, the actual final reference is to a text file
but the issue wasn't there.
Nor is it the only place; the one we shifted to in the message I saved
is the Ethernet header. Linux is full of protocol headers of all
flavors encoded as structs; many of those are suspect and may be
broken.
Tried to find an archive of this mailing list but wasn't successful.
The website lists three. Dejanews doesn't appear to do mailing lists;
findmail is out of order; and barnet appears to have nothing newer
than mid 1998.
unsubscribe: body of `unsubscribe linux-arm' to [EMAIL PROTECTED]
