Memtest - 4.22command-line memory testing utility |
|
||||||||||||||||
|
|||||||||||||||||
Feedback Summary:
| This Version: | |||||
| Overall Rating: | Not rated (0.0) | Features: | Not rated (0.0) | Support: | Not rated (0.0) |
| Ease of Use: | Not rated (0.0) | Quality / Stability: | Not rated (0.0) | Price: | Not rated (0.0) |
Key to Types of Feedback:
Reviews
Troubleshooting
Usage Tips
Developer Notes
Commentary
Featured Reviews
How to legally get memtest without paying for the download - Version: 4.21, 1/6/2008 03:58PM PST
(5 of 5 users found this comment useful)
skinlayersMost Recent Replies: View All 2 Replies
- Applejack = another FREE Memtest method
Where's the source? - Version: 4.21, 1/5/2008 07:04PM PST
(1 of 1 users found this comment useful)
Anubis IVMost Recent Replies: View All 3 Replies
- It's legal to charge a download fee (1 replies)
Purchase options - Version: 4.21, 9/11/2007 01:01PM PST
jimmy-d
The developer could charge a higher fee, and automatically email the latest revision the way SoftRAID does. That should address his concerns and not be a hassle for users wanting the latest version.
What you need to understand... - Version: 4.21, 8/27/2007 03:38PM PST
(1 of 1 users found this comment useful)
Don10So, even if you have paid the 1.39 charge to download a previous version, if you wish to download the latest version you must pay again, regardless of whether it is a major or minor update. When you 'purchase' the software you are given an expiring URL to your download. If you need to redownload there is that option to request a new URL, but this is just for the version previously downloaded, not the latest version. You get a password protected .dmg which you unlock with a password provided with the URL email.
Whilst I'm not sure that you can produce a demo of a CLI software, I would have thought the password-protected .dmg would be sufficient to allow paying supporters of the product to stay current with bug fixes etc, rather than have to pay each and every time - otherwise what incentive is there to support the developer? This needs to be made clear.
Most Recent Replies: View All 1 Replies
Tests all memory in Tiger 



- Version: 4.13, 8/16/2006 05:38AM PST
(2 of 2 users found this comment useful)
ellajI also agree with the other reviewers that memtest finds errors that Apple Hardware Test misses.
I have 5.5GByte installed on my G5 2*2GHz, and memtest found several errors, although AHT found nothing.
To identify the culprit was not an easy task, however, since the results obtained removing and putting back memory modules were not quite repeatable, even when running in single-user mode. Checking this much memory is also quite time-consuming, and the fan runs at top speed all the time...
I finally solved my problem by changing the order of the DIMMs, putting the largest memories in the first pair of banks, and smaller memories in the later banks.
As far as I know, the order shouldn't matter as long as the pairs are of equal size (and preferrably brand), so maybe it was just the fact that all modules were plugged out and then back in that did it, but now I can run memtest without errors with all memory installed.
Hope this can help someone.
Catches what others miss 



- Version: 4.12, 12/21/2005 07:21AM PST
(4 of 4 users found this comment useful)
Andy ShiekhCaught memory failures that other utilities missed.
identified bad DIMMs that Apple Hardware Test & Tech Tool Pro missed 



- Version: 4.12, 12/5/2005 10:16AM PST
(5 of 5 users found this comment useful)
doubleechowithout memtest, i'd have spent countless hours pursuing a non-existent software conflict as the root cause instead of what i know now to clearly be a previously undiagnosed hardware problem.
the author, tony scaminaci, was exceptionally responsive to a couple of simple questions from me. i particularly appreciated that as i ventured forward into new territory. kudos to tony.
unclear to me why apple hardware test and tech tool pro haven't incorporated the algorithms from memtest into their products since memtest found problems that AHT and TTP both missed.
the documentation recommends single user mode (unix command line prompt) for the most reliable and thorough test. i'm a big chicken when it comes to command line prompts (that's why i have a mac, of course!). but, this version runs fine from the desktop. memtest can only reserve about 2GB of memory at a time, but from the desktop several occurrences of memtest can be run at the same time. i found 2 bad DIMMs before i ever needed to simultaneously check all 8 DIMMs in my computer, but i did have 3 memtests running at the same time checking 6 DIMMs.
i also recommend using Rember, a close partner to memtest. it provides an easy way to quit all programs without having to quit them individually.
memtest is not without its deficiencies. as i mentioned, it will only test the first 2GB it can lock, and you don't get to pick which 2GB that is. so, if you have more than 2GB, you have to run more than one memtest at a time. would also be nice to be able to test the first 512MB or so of memory that the desktop uses before memtest even starts. the only way i could check that memory was to swap those DIMMs with known good DIMMs from earlier tests.
when memtest identifies a fault with a memory location, it doesn't correlate the error to a specific DIMM. so, i had to swap DIMMs in and out many times in a logic pattern and correlate the presence or absence of errors to the changes in DIMM configuration.
the documentation was sufficient, but perhaps a bit sparse for a novice like me. though i got the job done, i had many questions as i went along. most of the time i could infer the correct answer, but a couple of times i had to experiment a bit to figure it out.
don't read too much into my nitpicks. remember that memtest is free, but it solved problems that both AHT and TTP missed. that makes memtest invaluable, worthy of a five star rating for this stage of its development, and an important tool in every macintosh owner's box of tricks.
"dyld: incompatible cpu-subtype" on Tiger, not Panther - Version: 4.11, 8/20/2005 08:59AM PST
(5 of 5 users found this comment useful)
philip.kellerI wasn't able to figure out where this error is coming from. The project is built for the PowerPC, and with libraries statically linked. "otool -L" shows that the only dynmically linked libraries are /usr/lib/libmx.A.dylib (v. 92.0.0) and /usr/lib/libSystem.B.dylib (v. 88.0.0), so the problem must be in there somewhere. But where?
BTW, memtest did corroborate my suspicions about my RAM. A very, very useful program! As the documentation says, it's best to use memtest in single-user mode. That means typing one simple command instead of pointing and clicking. Most computer users should be able to survive it, if they just read the instructions (for a change).
A major benefit of the single-user mode approach is that your file system can remain read-only, and will not risk getting trashed by memory problems. This is no idle theorizing, but painful experience: I first started suspecting my RAM when my file system got badly corrupted and my only recourse was to erase and reinstall.
Most Recent Replies: View All 2 Replies
Installer Still Leaves Files Scattered - Version: 4.11, 7/31/2005 04:37AM PST
(4 of 4 users found this comment useful)
jamesladd
1) install Rember
2) Boot into single-user mode:
hold command-S on boot, then:
/sbin/fsck -fy
/sbin/mount -uw /
3) Once you have the hard drive mounted, run the memtest executable buried in Rember. If you installed Rember to /Applcation/Utilities the path would be:
/Applications/Utilities/Rember.app/Contents/Resources/memtest
For memtest switches look here:
memtest 4.21 user guide