News for hackers. Stuff that matters.
Re:Raspberry Pi microSD card performance comparison by rmu09 on Friday April 06, 2018 @03:03PM (#16776325)
I got 3 Samsung EVOs (from 2015 or so) that got stuck in a kind of "read only" mode after some months in the Pi. All seems normal for some time, but writes are not persisted. After some time linux gets confused and panics. After rebooting, FS is always in exact same state.
Re:Raspberry Pi microSD card performance comparison by cmurf on Friday April 06, 2018 @06:49PM (#16778170)
Same. I have a 32G Samsung EVO+ I use in a NUC with Linux that just went "read only" this week. 10 year warranty, and it's already setup for warranty replacement, which is sorta hilarious seeing as it cost $12, including the adapter (which itself is only warrantied for 1 year). I got 360 days out of the card, non-continuous use. But also interesting, they want the bad card returned to them - they're including a prepaid return shipping label.

The usage was just as a boot drive no user data: EFI FAT, ext4 /boot, Btrfs / using zstd and ssd_spread as the mount options.

FAT will mount ro or rw, but any writes fail

    [140718.615921] print_req_error: I/O error, dev mmcblk0, sector 2048
    [140718.615998] Buffer I/O error on dev mmcblk0p1, logical block 0, lost sync page write
Ext4 is similar but a lot more complaints from the mmc block driver, including an SDHCI REGISTER DUMP [1] which I can't make heads or tails of, but this is somewhat more revealing:

    [142132.340226] f28s.local kernel: mmc0: Card stuck in wrong state! mmcblk0 card_busy_detect status: 0xf00
Btrfs also has lots of complaints [2] but mainly because it never gives up trying to write. So cancel that and mount with '-o ro,norecovery' and everything is there. One thing you'll see in the Btrfs output, is "corrupt 10" which is not a current event, just a counter that I never reset. A few months ago I did have a data file (not fs metadata) become corrupt which Btrfs caught. I was able to reinstall the RPM that provided that file. So who knows how that corruption would manifest without being caught by the file system.

The blkdiscard command succeeds without error, but also doesn't actually do anything, all data is still there.

[1] https://drive.google.com/open?id=1Lbypuut21PreXnzHj9uC0x6lxK...

[2] https://drive.google.com/open?id=1340GQN29j8Ougtj_E0ey3A_RUJ...

Re:Raspberry Pi microSD card performance comparison by written on Friday April 06, 2018 @10:31PM (#16779085)
Sounds too cheap, are you sure it's not a fake?
Re:Raspberry Pi microSD card performance comparison by cmurf on Friday April 06, 2018 @11:01PM (#16779168)
Local reputable store, f3 had no complaints and is the claimed size. Samsung wanted a copy of the receipt and a photograph of the back of the card itself before replacing it so presumably they're satisfied it's legit. It might've been on sale for a few bucks off, I don't recall. Newegg has it for $15. shrug
Re:Raspberry Pi microSD card performance comparison by makomk on Friday April 06, 2018 @05:56PM (#16777814)
I had a cheap off-brand Chinese SD card do that a while back. Was very confusing. Surprised that reputable brands do the same thing though.
Re:Raspberry Pi microSD card performance comparison by thebeardedone on Saturday April 07, 2018 @03:35AM (#16779804)
I have been victim of 3 evos myself, i didnt get very far into the configuration process (db, transmission, etc) before them more or less being rendered useless. The ssd evos are great but the sd cards are simply horrible, tragically I saw them locally bundled with pi starter kits. I ended up buying a sandisk which has not had a single hiccup even after a little bit of abuse.