On 6/21/24 6:56 PM, Linus Torvalds wrote:
But I also know that pretty much *EVERY* time the SCSI layer has decided to start looking at some new piece of data, it turns out that "Oh, look, all those devices have only ever been tested with operating systems that did *NOT* look at that mode page or other thing, and surprise surprise - not being tested means that it's buggy".
We got the message and we will do what we can to prevent future regressions for USB devices. As has been mentioned earlier, there is evidence in sd_read_write_protect_flag() that SCSI devices may misbehave when querying a mode page. However, I was not familiar with that code and hence was not aware of the comments in that code. According to the git history, these comments were added before 2005, that is before I started reading the linux-scsi mailing list.
My argument is that things should be opt-in. If it wasn't needed for the previous 30 years go SCSI history, it sure as heck didn't suddenly become necessary today. So you literally NEVER DO THIS unless the system admin has explicitly enabled it. That's what opt-in means. And honestly, then the Android people can decide to opt in. Not random other victims.
>> What's the advantage of just enabling random new features that have no
real use case today? Put another way: why wasn't this an explicit opt-in from the get-go? And why can't we make that be the rule going forward for the *NEXT* time somebody introduces some random new mode page?
The new mode page has been introduced last year in SBC-5. UFS devices have a mix of SLC and TLC NAND internally and the new mode page allows
device vendors to reduce write amplification. This is important to UFS device vendors. I think that the new mode page is useful for all storage devices that have a mix of slow and fast storage internally and hence that it is also useful for some enterprise storage devices. This is why the new mode page is read by default. But as has been mentioned above, we have learned our lesson and will be much more careful in the future when adding code that modifies the access pattern of the sd driver for USB storage devices. Thanks, Bart.
