Export thread

I ran spinrite alpha release 4 on my 2012 macbook pro (intel) and it started off by running fast but then the screen went corrupted, so I filmed it on my cell phone and put it on a private youtube (links only). This is a used macbook from ebay, I owned it for about 5 months and used it about every night to read kindle and watch youtube/udemy. I’ve had no problems so far. 180GB SSD hard disk.

Also, I know the nnpt style new group would have been better, but I don’t have an account there, maybe someone can let the other news group know.

How did you create your boot media?

DanR said:

How did you create your boot media?

Click to expand...

Well I know that initdisk had something to with spin rite booting and when I tried it I only got a blank fat32 file system with nothing to boot no files at all, so then I used rufus 3.11 to create a freedos boot, is something wrong with using rufus?

coffeeprogrammer said:

Well I know that initdisk had something to with spin rite booting and when I tried it I only got a blank fat32 file system with nothing to boot no files at all, so then I used rufus 3.11 to create a freedos boot, is something wrong with using rufus?

Click to expand...

OK I see "freedos" argument, I will try and reformat with that BRB

coffeeprogrammer said:

OK I see "freedos" argument, I will try and reformat with that BRB

Click to expand...

Ok, I used InitDisk with "freedos" argument and it created the boot media and I tried again, but I still get the corrupted screen after it runs for a couple minutes

Well I feel like I should not be the only one with a 2012 macbook, it is a popular model MD101LL/A

coffeeprogrammer said:

Well I feel like I should not be the only one with a 2012 macbook, it is a popular model MD101LL/A

Click to expand...

I just tried on a thinkpad x220 tablet and it worked like a charm. 6:01 for 120GB SSD.

I don’t think is spinrite exactly on the macbook, I booted it again and left it at the command prompt and after a few minutes it pop into the corrupt screen, must just be freedos.

coffeeprogrammer said:

is something wrong with using rufus?

Click to expand...

Not necessarily. I have had problems with a Rufus generated boot drive, as have others. It worked on some of my machines but not on others. Other folks have also had mixed results with Rufus.

The InitDisk approach installs a plain vanilla FreeDos boot environment which may not have adequate graphics support for some systems.

I prefer the ReadSpeed approach, which installs a custom FreeDos environment., You might try that: https://forums.grc.com/threads/how-to-create-a-bootable-usb-thumb-drive.967/

I have been using my ReadSpeed created bootable thumb drive for all SpinRite development testing with no issues. Worth a try?

As I stated elsewhere: Steve is not looking at these forums for feedback. If you intend to test SpinRite alphas, and betas, you should be participating in the newsgroup and use the gitlab to open bug report tickets.

@coffeeprogrammer Found your post as I'm having the same problem with SR 6.1 prerelease on a 2012 MacBook Air with 256GB SSD. My screen goes "gray" not fuzzy, but almost exactly at the same point in SR level 2 run. I'm able to run ReadSpeed, benchmark the internal SSD with SR, and start the level 2 test. Three times the screen went blank only a few min into an estimated 9 min run. I'm using SR 6.1 Prerelease on a FreeDOS USB formatted by ReadSpeed on a Win10 machine. All of these downloaded today.

@saguaro Yes, I did want to help out in testing SR 6.1, I do look forward to Steve sharing about his process in developing it, but it became to difficult to continue to test because I start so many project and often there is no option to follow through the way Steve does. I gave my mac book to my mom because I am not not an apple fan, so I can't run a test today. I think Steve has a gitlab.com based testing forums, I am sure other users can refer you to those. I hope no one finds any more problems, I want to hear about Steve development on RTOS for SR 7!

@saguaro : Huh. That's really odd. We've had reports of SpinRite working on compatible macs... but nothing about screens blanking after a while. And I can't imagine what might be going on. Sorry!

Hello @Steve. NP. I was hopeful for SpinRite on mac, but it seems I'm outside the venn diagram of workable models. My 2012 MB Air gets gray screen and a 2015 MBP also in my house cannot boot from FreeDOS. It's interesting that I get the exact same result shown in @coffeeprogrammer 's video every time I try SR on the Air. Perhaps our Macs shared the same SSD, controller, or part of the firmware code. I'll give a 2018 intel-based mini a try, but that's almost certainly not going to work. I'd pass on my findings, but I'm not set up in the newsgroup and don't have the bandwidth available right now...

Still 6.1 is pretty darn impressive on my old wintel box!

Hi @Steve
Got here by doing a web search. I just purchased 6.1 after owning 6 for years; finally! It worked great and saved the ssd I was trying to salvage buried inside her imac 2012; really hard to get the components out of these. Mission accomplished.
Decided to run it on my 2012 Macbook pro, a laptop I have obviously been using for a long time. The most versatile system I have ever owned, having run linux mac and windows on it, now newer mac OS version using OpenCore. Drive is relatively new and working fine. This is maintenance.
Note: I believe 6.0 worked fine on this but I lost the installer and I wiped that USB stick because 6.1 seemed perfect. LOL.
What happens: It boots up fine and shows your splash screen. Hit a key nad it goes to the next screen no problem. I can hit escape and go to freedos, no problem. Launch spinrite.exe again no problem. What happens is when you hit any key and it goes into the new memory scanning portion where it displays a lot of things on the screen quickly. After a second screen displays a lot of little raster lines but is mostly grey. I think its this new function that causes the issue.

This is what system profiler says about this card. My guess is it can be reproduced on any system with this card mac or windows, but I haven't tried.

Chipset Model: Intel HD Graphics 4000
Type: GPU
Bus: Built-In
VRAM (Dynamic, Max): 1536 MB
Vendor: Intel
Device ID: 0x0166
Revision ID: 0x0009
Metal Family: Supported, Metal GPUFamily macOS 1
Displays:
Color LCD:
Display Type: LCD
Resolution: 1280 x 800
UI Looks like: 1280 x 800
Framebuffer Depth: 24-Bit Color (ARGB8888)
Main Display: Yes
Mirror: Off
Online: Yes
Automatically Adjust Brightness: Yes
Connection Type: Internal

I am sure if I talk to support I can recover my spinrite 6.0; however, would be absolutely great to help you figure this one out if you like.
Thanks for the great product and show over the years- I have no problem purchasing a few more licenses as payback ;-)

storkinsj said:

I am sure if I talk to support I can recover my spinrite 6.0; however, would be absolutely great to help you figure this one out if you like.

Click to expand...

Just start SpinRite from the command line as follows:

c:\>spinrite noramtest

This will skip the problematic (for you) ram test which your display does not seem to like.

storkinsj said:

I am sure if I talk to support I can recover my spinrite 6.0; however, would be absolutely great to help you figure this one out if you like.

Click to expand...

If you have your licence key, go to https://www.grc.com/cs/customerservice.htm, put in your licence key. When it gives you a download link, DO NOT CLICK ON IT, copy the link, and edit the download link, replacing "spinrite.exe" with "sr60.exe". This will let you re-download a copy of v6.0

Some updates: First thanks for the tips. In sections:
Spinrite 6.0:
I found my spinrite 6.0 iso image on a backup of an old computer and was able to try it. Apparently tales of running it before with no problems have been vastly exaggerated. Soon after it launches I have no control over the keyboard so I am not sure how I actually ran it before.
Spinrite 6.1:
Before seeing your good suggestion here I decided that Steve would have been likely to implement spinrite /h and it at least trapped me into the paramaters screen. The initial screen did not cover noramtest but the exit screen did list it without the same level of documentation. I did try spinrite auto level 2 with no luck and the same results.
Spinrite noramtest got me much further. I was able to select the drive and the level and start spinrite. After a while however, the screen became corrupted again.
I tried to put spinrite on a different screen instead; I used the "screen saver" screen where the box moves around. I got over 20% complete before maybe it switched back to the main screen with the progress and the o's that seems like a sector map (but cant be because it always displays the same number). After a while that seems to cause the screen corruption.

I was hoping to turn on logging so that I could see how long it went and if it continued after the screen was corrupted but I can not exit out of that settings screen for some reason.

So I believe: Much further along but still blocked. There may be some way of operating that gets me all the way but there are definitely some problems. I am guessing they are not specific computer hardware's issues or we would not all be experiencing the same problem.

Of course I'm selfish in this regard but I think it's worth solving down the line. A tiny bit of "research" using AI came up with these numbers. Try to find a mac that can run sequoia that has dvd/media, ethernet sd card port, thunderbolt, usb 3, magsafe... and a headphone jack not to mention the really well done backlit keyboard. It's not a wonder. I have upgraded my hd and my ram (maxed at 16Gb). If you asked me if you should get one on ebay, I would probably say yes.

Known Facts About the Model

Model: MacBookPro9,2 (13" Mid-2012)
This was the last non-Retina, upgradeable MacBook Pro (removable RAM + 2.5" SATA + optical bay).
This is the single most sold MacBook Pro model of its generation (volume sales between 2012–2015, especially into education and enterprise).
Original estimated unit shipments (industry analysts, SEC filings by component suppliers, teardown-linked volume signals):
Roughly 10–15 million units sold globally across its market lifespan.
Attrition / Survival Curve (Typical Laptop Decay)

Consumer + education laptops follow a well-studied survival curve:
~80% still in some form of use after 3 years
~55–60% after 5 years
~15–25% after 10 years
~5–10% after 12–14 years (if model is not special)
But this model is special:
Upgrade-friendly → many users kept them alive with SSD + 16 GB RAM
Strong appeal in developing regions, right-to-repair crowd, Linux community
Because of that, it retains a significantly higher survival rate than sealed models from 2016+.
Adjusted survival estimate for this specific model at ~13 years:
→ ~12–18% remaining active.

So the Math

If ~12M were sold and ~12–18% remain:
12,000,000 units × 0.12 = 1.44M
12,000,000 units × 0.18 = 2.16M

Likely active worldwide today: ~1.4–2.2 million units.

However:

Many are used as secondary / offline systems
Many run Linux instead of macOS
Many are sitting in closets but bootable

If we define “in regular daily or weekly use”, a good working range is:

~400,000 to 1,200,000 still being actively used worldwide.

storkinsj said:

Decided to run it on my 2012 Macbook pro, a laptop I have obviously been using for a long time.

Click to expand...

….

storkinsj said:

What happens is when you hit any key and it goes into the new memory scanning portion where it displays a lot of things on the screen quickly. After a second screen displays a lot of little raster lines but is mostly grey. I think it’s this new function that causes the issue.

Click to expand...

I have the same MacBook Pro and the same thing happens on mine. You have to remember that Macs aren’t entirely PC compatible. I had better luck creating an external MacOS boot disk and running SpinRite from there under VirtualBox.

Instructions for that are here:

https://forums.grc.com/threads/how-to-run-spinrite-on-a-uefi-only-machine-part-1-of-5.1613/

Thanks for the information!
A small correction: Intel macs really are PC's. Loaded with the right drivers (boot camp gives a perfect set) you can run Windows 10 flawlessly, and of course windows 11 with a registry tweak. UEFI and Bios based bootups are used depending on the version. The question is how many PC's is FreeDOS compatible. The fix to allow multi-HID keyboards in Spinrite 6.1 is GREATLY appreciated.

Question @Scott : You said creating an external macos boot disk. I guess this frees up the drive we are trying to maintain. I found my old spinrite VM in that backup I mentioned so I could sort of do things that way.
Do you think I can use target mode and thunderbolt, USB, or (I doubt it) firewire to do the repair? Obviously I would set virtual box or fusion to do direct drive access on the drive.

I am REALLY close to getting it working. Like I said it goes through 20% of the drive and then goes into that screen problem. I would rather leave notes here inperpetuity for the google search landers on how to get it working without a bunch of extra effort, and we are really almost there.

I was thinking it could be related to the screen "mode" too (rows and columns of text). I was thinking of grabbing the DOS mode command and dropping it onto freedos to change the screen resolution to a better tested one. This is a 13 inch screen so it's possible it got less testing time.

storkinsj said:

A small correction: Intel macs really are PC's

Click to expand...

No, actually they're not. They're a hybrid device. Their BIOS emulation is flawed, as you have seen. So I guess you can stick with, they attempt to be a "real PC" and succeed fairly well for specific use cases, but they're not actually a real PC for *ALL* use cases, one of which is SpinRite. For UEFI support, which is what Windows will use if available, I suspect they are much more compatible, as you noted.

However, note that there are many supposedly 'compatible' PCs that
have BIOS quirks that interfere with predictable FreeDOS and SpinRite
access to drives to be tested - it's an ad hoc universe out there.

So just using a "PC" doesn't guarantee compatibility.

And compatibility with what? PC? XT? AT? How far back in 'pristine'
IBM-compatibility do we have to go, and then precede USB and SATA?

So the word 'compatible' is a hunt-and-find thingy that basically means
"it works" or not.

Some older motherboards have specific BIOS "defects", such as a bug
with USB access on older systems or issues with data transfer
initialization, that can cause data corruption or prevent SpinRite from
proceeding with a scan, forcing the user to use a different computer
or a virtual machine setup.

I've got one computer that considers all USB drives to be the same
size as the first-discovered USB drive - probably not what I want to
start re-writing on.

Then there's the Roger Anomaly "... that all data written during a
maximum size multi-sector transfer, which immediately follows a
maximum size multi-sector read, is "shifted down" or "skewed" by
exactly 32-bits, or four bytes. 32-bits of zeroes are written at the
beginning of the transfer and the last 32-bits of each sector are shifted
into the first 32-bits of the following sector. The final 32-bits of the last
sector fall off the end and are lost forever ..." see
https://www.grc.com/sr/kb/badbios.htm Supposedly, SpinRite 6.1 is
immune.

Regardless, all we can do is try everything, then move to another
computer and try everything again.

Luckily, old used BIOS PCs are cheap and even freely available, and
so long as it has USB2 and SATA2 or SATA3, it will probably be a
useful SpinRite testbed, a keeper.

I even get serviceability from diskette booting, USB1, PATA, and
adapters - whatever works, works.

We all probably ought to keep at least one old PC hanging around.

- - - - -

That said, there are a variety of ways to simplify SpinRite to maybe
avoid BIOS quirks:

Change CONFIG.SYS to contain only:

BUFFERS=98

Load SpinRite 6.1.8961 R4 with the following command line:

SPINRITE.EXE NORAMTEST BLANKING 0 DYNASTAT 0 NOREWRITE SKIPVERIFY FORCEBIOS GRAPHIC NOBLANKING QUIET ROCFG VIDEO

Have I found everything?

SpinRite 6.1 command line options are nicely summarized here:
https://gitlab.com/GRC-Community/spinrite-6.1-wiki/-/wikis/Command-Line
- thanks to @ColbyBouma.

Thank you- I was going to say something similar but I will follow your suggestions and get back to you. Also interesting that most non apple computers that are intel based have also switched to UEFI from bios.

Ironically I might add, I did try forcebios earlier and got a warning from spinrite that using the actual bios will reduce its capabilities. So the bios-love seems a bit unwarranted. Nonetheless thanks to everyone of any opinion for your help in this thread.

storkinsj said:

Also interesting that most non apple computers that are intel based have also switched to UEFI from bios.

Click to expand...

Not really. It's an Intel initiative. Since 2020 or so they have declared EOL for the BIOS. I think there is info about that in other posts on the site. But this Reddit most may be the source of some of such info: intel_drops_legacy_bios_support_in_2020

storkinsj said:

Ironically I might add, I did try forcebios earlier and got a warning from spinrite that using the actual bios will reduce its capabilities. So the bios-love seems a bit unwarranted. Nonetheless thanks to everyone of any opinion for your help in this thread.

Click to expand...

The warning from SR 6.1 about BIOS is that using "forcebios" limits SR to reading/writing the drives via BIOS calls which are significantly slower than using the supplied PATA or SATA drivers.

For other devices, e.g. M.2 or USB drives, going via the BIOS is the only option, but for PATA or SATA drives, SR6.1 now contains native drivers which speed things up considerably.

"... I did try forcebios earlier and got a warning from spinrite that
using the actual bios will reduce its capabilities. So the bios-love seems
a bit unwarranted ..."

The 'warning' is not about danger, but about comparative speed.

@Steve Gibson is justifiably proud of the comparative speed of
SpinRite 6.1 IDE/ATA/AHCI drivers, an important upgrade
considering the growth in size of hard disk drives compared to 20
years ago - a mere 'fix' would not feel like much of an improvement
if speed did not also upgrade to match the larger drives.

However, if BIOS is the only way to get to a drive, then BIOS on.

It's not "BIOS love", per se, it BIOS DOS because that's what was
available from SpinRite v1 up to v6, which v6.1 is a 'fix' for.

Updating BIOS-DOS-based v6 to a non-BIOS non-DOS environment
would be a mismatch for SpinRite v6 users wanting to eliminate
known errors.

SpinRite 6.1 eliminates SpinRite 6 errors, and adds speed and other
savvy - a bonus.

SpinRite 7 will move to a different environment, and is not a 'patch'
for prior SpinRite misbehavior.

Does that help folks understand SpinRite 6.1?

PHolder said:

Not really. It's an Intel initiative. Since 2020 or so they have declared EOL for the BIOS. I think there is info about that in other posts on the site. But this Reddit most may be the source of some of such info: intel_drops_legacy_bios_support_in_2020

Click to expand...

You just agreed with me.

Here's where I am and I think I have to give in at this point.
-All scenarios, I get the blank/corrupt screen even with all those options.
-In fact I can get that if I just leave the system at the free-dos prompt with no spin-rite. Leave it long enough and the screen gets corrupted.
-It could be working. I ensured logging is on with long sentences, tried viewing the spinrite log but it only logs if it sees a problem. I dont see a way to turn that up to a true activity log. And there are gigabytes of free space on the usb drive.

But it could be working. It could only be a screen issue.

storkinsj said:

So the bios-love seems a bit unwarranted.

Click to expand...

The BIOS is to SpinRite what Windows is to say MS Word. The BIOS is a part of the Operating System of [Free]DOS, and it supplies many capabilities, such as reading the disks (but SpinRite will use its own drivers if it has them), outputting to the screen, reading the keyboard, and generally providing the memory location of the devices installed in the system (such as video cards, disk controllers, etc.) For SpinRite 6.1 and earlier it CANNOT operate without a working, and generally 100% compatible with expectations, BIOS.

storkinsj said:

agreed with me

Click to expand...

I agree that the BIOS is becoming legacy, yes, but I don't agree that that is notable in any particular way, which is what I said.

storkinsj said:

Here's where I am and I think I have to give in at this point.
-All scenarios, I get the blank/corrupt screen even with all those options.
-In fact I can get that if I just leave the system at the free-dos prompt with no spin-rite. Leave it long enough and the screen gets corrupted.
-It could be working. I ensured logging is on with long sentences, tried viewing the spinrite log but it only logs if it sees a problem. I dont see a way to turn that up to a true activity log. And there are gigabytes of free space on the usb drive.

But it could be working. It could only be a screen issue.

Click to expand...

Again, I have had SR work against this MacBook Pro using both VirtualBox and QEMU when booting from an external drive running Linux or MacOS

Ultimately I was unable to get this working all the way through to my knowledge. Logging is too weak to suffice for understanding if it worked; maybe a good feature request.

FWIW I do have an old dell somewhere I kept exactly for this. It unfortunately is across an ocean but it is a great idea.