Linux @lemmy.ml petsoi @discuss.tchncs.de 4w ago

Linux 6.13 Staging Clears Out 107k Lines Of Code From Old & Unmaintained Drivers

31 comments

Deleting 107k lines of code is so much more based than adding 107k lines
- It really is. I always make a note to point out how much code is removed in PRs I review
  
  Code is a liability.
  
  You could probably build a tool that assesses the risk of any given PR based on this and several other signals. PRs with enough risk should require justification and sign off.
Wait I was using one of those drivers!
- Heads will roll if my LS-120 drive stops working!
  
  Is there somewhere to get these drivers if you want to run really old hardware?
  
  I was devastated when I couldn't use my floppy drives anymore.
- o well
https://xkcd.com/1172/
- this is so supremely cursed
Yay bloat be gone \O/
Why clear them out if they still work and are useful? Seems like a backwards step. What's that phrase that people throw about:sometimes things are just done and don't need changing.
- Unmaintained code in the kernel is really bad due to possible vulnerabilities. If you want to keep it, it must be maintained.
- As the kernel moves on changes could be introduced that make them difficult to compile with the new kernel. Unmaintained doesn't only mean not adding new features, it means keeping up with the rest of the code.
  
  From another article
  
  The GDM724x is removed for supporting the GCT GDM724x LTE chip based USB modem devices. This driver was merged back in 2013 but is being removed now as the driver isn't being maintained and yields a maintenance workload, the manufacturer GCT doesn't respond to any emails/support, there doesn't appear to be any of the said chips easily available for purchase, there is not any hardware documentation available, and no apparent usage of this driver remaining in the Linux community. Removing the driver clears out 3.6k lines of code and lowers the maintenance burden for other kernel developers.
  
  There was also a vulnerability discovered in July linked to this driver.
  
  So yeah I understand that they chose to remove some drivers from the kernel.
  
  But that's true of all code in the kernel. If any change can break something then all broken bits will need fixing. Why not remove all drivers in case an update breaks them. Things can't be preemptively fixed before breaking changes are made. A driver can be complete and only need updating if someone else breaks stuff, so leave it alone until then and only remove it I'd no one comes to fix it.
  
  Removing functionality just in case is daft.
- the less code the better
  
  No. The less code for a given set of functionality the better... often. Removing functionality just to reduce code is daft. Otherwise stop adding any features. Remove all features of the kernel until machines only just boot. Lot less code!
- Have you SEEN the base install size for Windows 11?
- because nobody is updating them and the one person that did before was seemingly the only user. Nobody could find any evidence of it being used. When was the last time you ever heard about fieldbus?
  
  the article mentions support for a another interface is also being added, for lab equipment that actually does still get used.
This seems like a bad idea... What about people using hardware that needs those
- This is from the article: „If there are any genuine users of these drivers remaining that are still running an upstream kernel, the drivers can always be reverted / merged back but otherwise they are gone without anyone maintaining them.“
  
  All well and good for people who know how to do that
  
  A lot of users won't even know what a kernel is let alone why their printer has stopped working or that they need to raise a GitHub issue
This normal for staging merge?

31 comments