Weekly edition	Kernel	Security	Distributions	Contact Us	Search
Archives	Calendar	Subscribe	Write for LWN	LWN.net FAQ	Sponsors

GPLv3 glibc

Posted Jan 3, 2008 2:19 UTC (Thu) by stevenj (guest, #421)
Parent article: Wild predictions for 2008

A GPLv3-licensed glibc would have to be user-replaceable, which could be problematic on locked-down devices.

Clarification: it would have to be user-replaceable if it is replaceable by the vendor. GPLv3 code does not have to be replaceable on a truly "locked-down" device, one which is not designed to be modifiable by anyone.

(Log in to post comments)

GPLv3 glibc

Posted Jan 3, 2008 2:50 UTC (Thu) by jimparis (subscriber, #38647) [Link]

That's interesting, I hadn't realized that. For the benefit of others, the relevant clause:

If you convey an object code work under this section in, or with, or specifically for use in, a User Product, and the conveying occurs as part of a transaction in which the right of possession and use of the User Product is transferred to the recipient in perpetuity or for a fixed term (regardless of how the transaction is characterized), the Corresponding Source conveyed under this section must be accompanied by the Installation Information. But this requirement does not apply if neither you nor any third party retains the ability to install modified object code on the User Product (for example, the work has been installed in ROM).

GPLv3 glibc

Posted Jan 4, 2008 3:15 UTC (Fri) by dmag (subscriber, #17775) [Link]

Nowadays, Flash is cheaper than ROM for all but the smallest microcontrollers (like 8 bit
size). GlibC is huge, so it doesn't make sense to put it in ROM. If it's in flash, they're
going to leave themselves a way to write it (even if it's just thru JTAG). Therefore it's
unlikely anyone distributes a completely non-writable GlibC (or even Linux kernel for that
matter).

This is already an issue with LGPL v2

Posted Jan 3, 2008 4:43 UTC (Thu) by JoeBuck (subscriber, #2330) [Link]

Even with LGPL v2 (the existing glibc license) there has to be a way to replace the library. For that reason, there have long been alternative C libraries for embedded systems, for example newlib, first developed by Cygnus.

This is already an issue with LGPL v2

Posted Jan 3, 2008 5:46 UTC (Thu) by stevenj (guest, #421) [Link]

Just as for the GPLv2, it is debatable whether this clause has any teeth when there is a hardware-signature mechanism that prevents replacement code from working. Certainly, the FSF was worried enough about this "Tivoization" loophole to explicitly address it in v3 of the license.