Tiny Core Linux
Off-Topic => Off-Topic - Tiny Core Lounge => Topic started by: gadget42 on February 08, 2024, 10:08:24 AM
-
counterfeits and fakes continue, buyers be aware and do due-diligence always
https://arstechnica.com/gadgets/2024/02/rejected-chips-hidden-microsd-cards-plague-the-usb-stick-market/
-
Hi gadget42
Last summer I purchased an 8 Gig SD card for my camera
at Best Buy. I used the f3 (Fight Flash Fraud) utilities on it:
https://fight-flash-fraud.readthedocs.io/en/latest/introduction.html
and it failed, so I returned it. Last December I ordered a couple
of 4 Gig cards from Amazon, and they passed.
I downloaded version 8.0 from here:
https://github.com/AltraMayor/f3/tags
Then:
tce-load -wi compiletc
unzip f3-8.0.zip
cd unzip f3-8.0
make
Then to test a card:
mount /dev/sdf1
./f3write /mnt/sdf1
./f3read /mnt/sdf1I think the write process took about 10 minutes.
I don't recall how long the read process took, but
I'm certain it was much quicker.
-
more links for future visitors convenience:
https://fight-flash-fraud.readthedocs.io/en/latest/index.html
https://oss.digirati.com.br/f3/
https://fightflashfraud.wordpress.com/
https://sosfakeflash.wordpress.com/
and one for the microsoft windows people:
https://www.grc.com/validrive.htm
@Rich, thank you for your wonderful post!
-
I can very recommend ThioJoe short video regarding this scams drives.
And faster program that not scans the hole drive, sadly I think it's only for Windows.
Maybe some one can write similar program for the other platforms.
https://youtu.be/xMgEHy1A9QA (https://youtu.be/xMgEHy1A9QA)
-
Hi patrikg
... And faster program that not scans the hole drive, ...
As you saw in the video, these scammers create their own firmware
to make these devices look like something they are not. They already
create fake directory entries to make it look like files were saved. I'm
sure they are capable of figuring out ways to evade partial size tests
like ValiDrive which states it does a 576 region spot check. There's
nothing to stop them from keeping track of the most recently written
locations so they are available for reading back when requested.
Testing the whole device is the only way to be certain. Leaving untested
holes just leaves openings for the scammers to eventually exploit.
-
HI Rish
Truth be told, where there is money to be made, there are always crooks around.
-
mostly i pointed to ValiDrive because that webpage is fairly informative and gives an account of recent purchases of fraudulent drives.
(i haven't tried it as i don't do windows but i was using Steve Gibson's stuff decades ago when i did do microsoft)
i agree with Rich that the description does say "576-region spot-check" near the top of the page
further down on that webpage it says "ValiDrive performs a quick, random-sequence spot-check across the drive's entire declared storage space. At every location it verifies the successful storage and retrieval of random (unspoofable) test data."
actually this additional webpage gives more details:
https://www.grc.com/validrive/ui-details.htm
someday i may grab some old spinning rust and load up windows to play around with ValiDrive but it's not likely in the short-term
ymmv
-
just ran f3 on a 2GB flash media so i could do a paste:
somebody@somewhere:~$ sudo f3probe --destructive --time-ops /dev/sdd
F3 probe 7.1
Copyright (C) 2010 Digirati Internet LTDA.
This is free software; see the source for copying conditions.
WARNING: Probing normally takes from a few seconds to 15 minutes, but it can take longer. Please be patient.
Good news: The device `/dev/sdd' is the real thing
Device geometry:
*Usable* size: 1.83 GB (3842048 blocks)
Announced size: 1.83 GB (3842048 blocks)
Module: 2.00 GB (2^31 Bytes)
Approximate cache size: 0.00 Byte (0 blocks), need-reset=no
Physical block size: 512.00 Byte (2^9 Bytes)
Probe time: 2'28"
Operation: total time / count = avg time
Read: 1.12s / 4747 = 236us
Write: 2'27" / 1744961 = 84us
Reset: 0us / 1 = 0us
somebody@somewhere:~$
-
Hi gadget42
Just wanted to point out, f3probe does just that, it probes.
It does not write and read all locations.
There's also this:
... I’ve marked f3probe as experimental for now because this model has not been battle proven. ...
Found here:
https://fight-flash-fraud.readthedocs.io/en/latest/usage.html
-
just ran f3write & f3read on a 2GB flash media so i could do a paste:
somebody@somewhere:~$ f3write /media/somebody/2gb-minimee
F3 write 7.1
Copyright (C) 2010 Digirati Internet LTDA.
This is free software; see the source for copying conditions.
Free space: 1.83 GB
Creating file 1.h2w ... OK!
Creating file 2.h2w ... OK!
Free space: 0.00 Byte
Average writing speed: 5.35 MB/s
somebody@somewhere:~$ f3read /media/somebody/2gb-minimee
F3 read 7.1
Copyright (C) 2010 Digirati Internet LTDA.
This is free software; see the source for copying conditions.
SECTORS ok/corrupted/changed/overwritten
Validating file 1.h2w ... 2097152/ 0/ 0/ 0
Validating file 2.h2w ... 1735320/ 0/ 0/ 0
Data OK: 1.83 GB (3832472 sectors)
Data LOST: 0.00 Byte (0 sectors)
Corrupted: 0.00 Byte (0 sectors)
Slightly changed: 0.00 Byte (0 sectors)
Overwritten: 0.00 Byte (0 sectors)
Average reading speed: 16.46 MB/s
somebody@somewhere:~$f3probe done in my earlier post
-
ValidDrive and its "spot-check" is actually a reasonably "quick" test 90+% of the time -- in a few instances it crawled as the fakeware on the bogus flash (eg: "Shannanzi USB3.0 512GB") couldn't make heads or tails of the addresses that were thrown at it and choked. Out of three known "bogus" flash chips we tested (512GB China uSD - three different "brands" -- or "labeling" on the chips at least) all three were detected as fakes, their "real" size was properly detected (not that it matters as the firmware makes even the real physical size unreliable at best) and the amount of time it took for each averaged less than a minute (compared to byte-for-byte testing, that's nothing!)
Truth be told, where there is money to be made, there are always crooks around.
So sad... but so true.
-
this was a previous forum entry regarding counterfeits, fakes, and frauds:
https://forum.tinycorelinux.net/index.php/topic,26583.msg171356.html#msg171356
"brands" would relate to "resellers" and would be an unverifiable number in any case.
just try to get a valid/official chain-of-custody AND certifications-of-authenticity(for all the assemblies/components/processes) on even _one_ product!
the whole planet is awash with fake stuff!
for example, how do you KNOW each and every specified component at the motherboard level is authentic/real on a TALOSII that you just paid $10K for?
https://www.raptorcs.com/TALOSII/
what about all the components in the controls on civilian AND military rockets/missiles/biological-and-chemical-release-systems/weapons-of-mass-destruction-and-murder?
https://www.youtube.com/watch?v=LQUXuQ6Zd9w
referencing TALOS reminded of this unboxing commentary:
https://tenfourfox.blogspot.com/2018/04/unboxing-talos-ii-its-here.html
-
Split topic: Counterfeit and fake flash devices - let's fight back! (http://forum.tinycorelinux.net/index.php/topic,26778.0.html)
-
Hi!
Thanks for warning about the problem anyone may face. I see many people take participation in the struggle with fake flash. They write articles, write sophisticated code, create and maintain repositories, produce issues and pull requests... But does the problem really worth so much efforts? Isn't it enough to fill the drive with any non-repetitive sequence and then verify it? Why so much hype and movements? Just as the mean to attract as much attention as possible? But I guess anyone who will once face the problem will be aware enough :-) Others will continue spending their still lives, not bad too.
-
@jazzbiker: In the past two months, we've ordered a few dozen 512GB flash based devices (SD cards, MicroSDs, USB sticks, etc.) from numerous vendors. Our local hardware store (55+ minute drive... and I call it local) sells these cards legitimately for $35 and the quality is reasonably solid. Some of these "fake" vendors will go as low as $11 USD for a 512GB, I'm guessing with the mentality of "...stealing $11 is better than trying to get $30 and having fewer victims!" so even though there's a slight chance of financial loss, I've paid $11 through $28 for many different China brands and every single one of them is a fake. In fact, only one China chip was what it claimed to be... but it claimed to be 400GB.
400GB isn't "a thing."
More than likely, it was SUPPOSED to be a 512GB but the die likely cut quite a few rejects where 400 of the 512 was usable... so technically they're defective but they passed testing!
Now, to make fun of the problem, these same "fake" vendors are using naive shoppers as their newest victims by hacking old 1GB and 2GB SD cards and selling them as 128MB, 256MB and 512MB... most people don't "realize" that "MB" hasn't been manufactured in YEARS and it's not the size you're after, but these MB cards are selling like hotcakes because they have the right NUMBERS ("...my son told me to buy a 256...") and the hacks are LEGAL (if you buy a 512MB card --- you're GETTING 512MB!) so resellers, banks, credit card companies, etc. aren't as willing to approve charge-backs/refunds as you end up getting exactly what you purchased, even though you just paid $25 for a 256 MEGAbyte SD card!!!
does the problem really worth so much efforts?
Sometimes.
Isn't it enough to fill the drive with any non-repetitive sequence and then verify it?
To fill an SD card of 512GB at ~25MB/s... I'm guessing that's about six hours of just writing to the card, another 3-4 hours verifying the data ~40MB/s... all for a $35 card that I know where I can get legitimate ones... but our clients don't. Their friends and family... they probably do not as well. If I educate just ONE client, there's a possible impact of 20 or more people. If all of them were "informed consumers" that's potentially HUNDREDS of dollars saved from theft.
Most clones/fakes start at 128GB these days, but there are plenty of the smaller ones still in circulation. Plus, there's the MB issue, too.
We hunt down the bogus "brands" and where they can be purchased from and we find legitimate ones in the process. Sharing this information doesn't put the bad guys out of business, but if enough people know how to look for problems, that's $XXX.xx the bad guys aren't making!
-
Hi jazzbiker
... Isn't it enough to fill the drive with any non-repetitive sequence and then verify it? ...
To the best of my knowledge, yes it is. In my opinion, it's
also the only way to be absolutely certain there isn't any
slight of hand (funny business) going on.
My reply #1 does exactly that.
... Why so much hype and movements? ...
If there is any hype, it's probably about time and trying to
speed up the process by doing selective testing. My 4 Gig
SD cards took about 10 minutes to fill with data. If you
extrapolate that to a 64 Gig card it now takes 160 minutes.
If I have to let my computer work a while to confirm my SD
card can be trusted to save all of my data, it's worth it to me.
I often use my camera to document various steps in a project.
Since these steps often involve removing or adding material
to a part, those pictures can't be recreated if they are lost
due to a "faulty" memory card.
-
Hi Rich!
Hi jazzbiker
... Isn't it enough to fill the drive with any non-repetitive sequence and then verify it? ...
To the best of my knowledge, yes it is. In my opinion, it's
also the only way to be absolutely certain there isn't any
slight of hand (funny business) going on.
My reply #1 does exactly that.
I know that Your programming skills are based on the solid digital and hardware background. So You definitely know how flash memory works behind the scenes. I mean an internal block organization, when writing the block includes flashing (clearing) the whole block and then writing row after row where each row consists of N bytes until the whole block will be written. Such things can not be fooled around I believe. (Am I wrong?) The time of the block writing is Tclear + X * Trow. The block size is usually rather big, something of 256K or bigger in size.
What's Your opinion, can the user get the block size of the chips making the flash drive? It's about possible optimizations.
If about f3... Probably its code is perfect, but more than 1K LOC, Qt, Swift, Docker, ... holy shit, welcome to the Bloatshire!
Don't trying to say that someone should not use this package, if it does the job successfully it's all You need.
-
Hi CentralWare!
I didn't know that the plague is so widespread :-( I've never seen not a single fake flash, for me it is the story from the "internet", while You come across day by day :-(
I think as the first step You may promote f3 package among You clients, at least.
-
Hi jazzbiker
There's a controller that sits between you and the actual
memory chips. That controller contains firmware and is
responsible for determining which chips and when to
perform the actual write. The controller also performs
wear leveling by remapping heavily used locations. I
suspect it may also include some RAM to buffer the
data before it needs to be committed to flash. So it
would not be easy determining block size.
... If about f3... Probably its code is perfect, but more than 1K LOC, Qt, Swift, Docker, ... holy shit, welcome to the Bloatshire! ...
That's a bunch of optional stuff. All I compiled was
f3read and f3write, 51K and 50K respectively, and
that's not stripped.
-
was reviewing some of Michael Horowitz webpages and was at:
https://defensivecomputingchecklist.com/extracredit.php
and on that page he mentions this interesting device:
https://usbkill.com/
definitely thought-provoking
-
more regarding the fake stuff and the troubles(and wasted time) caused by it:
see this thread:
https://forum.tinycorelinux.net/index.php/topic,27348.0.html
and i did post a note for future thread visitors to steer those interested back to this earlier discussion:
https://forum.tinycorelinux.net/index.php/topic,27348.msg176088.html#msg176088
-
after reviewing this thread i wanted to add an additional link/article/comments regarding an earlier post mentioning a "usbkill"
https://www.schneier.com/blog/archives/2016/09/usb_kill_stick.html
yes, i know it is old news but it is still a present-day concern to anyone using any electronics and/or their associated peripherals.
-
stumbled across this webpage while researching something else and wanted to include it on our forum/threads/webpages:
https://docs.armbian.com/User-Guide_Getting-Started/#preparing-sd-card
definitely good to review the whole webpage but here is a quick snip:Preparing SD card:
Important note: Make sure you use a good, reliable and fast SD card. If you encounter boot or stability troubles in over 95 percent of the time it is either insufficient power supply or related to SD card (bad card, bad card reader, something went wrong when burning the image, card too slow to boot – ‘Class 10’ highly recommended!). Armbian can simply not run on unreliable hardware so checking your SD card with either F3 or H2testw is mandatory if you run in problems. Since counterfeit SD cards are still an issue checking with F3/H2testw directly after purchase is highly recommended.