Hank M writes:
Three weeks ago I bought via the Premium Apple reseller Amac your advised iMac, see below. Since then I have a lot of crashes in Lightroom 6.5 a day. Black screen and restart of the Mac.
I asked for Apple Care, Adobe, Queen Lightroom for advice and send them my crash reports. Everybody think the problem is RAM related, in other words the RAM of 64GB you have advised. The QE of Amac thinks it is because the OWC RAM is made for 64bits while the iMac is 32 bits.
Though the best way to find out is to replace the 64GB RAM for 32GB “Apple” RAM, 32GB OCW RAM did not work either. Because I paid a lot of money 800 Euro for the OCW RAM for a fast working Lightroom I expected it to work. It cost me a lot of time to stop every time with my work and correspond with Adobe and others.
I trust you research this phenomena and that you warrant all your customers and readers that OCW 64GB does not work with Adobe Photoshop Lightroom. Also Google Chrome does not load sometimes and I have got glitches on my screen when I play a video.
As I stated before, I thought it could be wise to change the RAM from the OCW 64GB to OCW 32GB. That did not help.
Then I changed OWC 64GB to Crucial 32 GB that helped BUT other things occur. Everything became SLOW, Google Chrome did not load always.
Also I change again with the OCW modules. Much faster but crashing again.
What I do before when my Lightroom 6.5.1 crashes (suddenly stops and the Mac makes a restart), are:
1. Zooming using the F-key in the Develop Module.
2. Deleting spots. This I do very frequent because I am busy with scanning my thousands of Middle Format slides with a Nikon D810 and a Macro Lens. The results are fantastic. Better than Drum Scanning. There are lot of dust spots to remove. After I had removed 30 spots my screen goes black. LR Crashes.
Most of the time this happens with this kind of adjustments in the Development Module of LR.
... I am sorry to tell that after I updated OS X to 10.11.5 and busy with the Spotting Tool, after 6 times LR crashed AGAIN!
DIGLLOYD: Hank M is very frustrated, and rightly so. Not just because of the issues, but because of the bad advice from everyone from which he’s asked advice. The message is not entirely clear, but I’ve gone back and forth with him on getting more details.
I’ve used Photoshop with 64GB memory on the MPG late 2015 iMac 5K for many months now in demanding work—it has been 'bulletproof'. As a check, I loaded up the latest Adobe Lightroom on the MPG iMac 5K with 64GB OWC memory, imported 880 raw files totalling 48GB, exported all of them to JPEG, then spotted 30 times without any issues. Whatever the issue, it is not happening readily for the MPG system.
There is a lot of bad advice out there handed out in ignorance, the graping-at-straws approach, since few people are capable of saying “I don’t know” (the 64 bits vs 32 bits comment foisted on Hank M is just an expression of incompetence). Having a degree in computer science, and having been a professional software engineer for 25 years writing interrupt-driven code, multithreaded server code, device drivers, C, C++, java and assembly, I know a 'little bit' about crashes and debugging. It is not an answer to take a fairly vague problem and blame it on “bad memory”—at least not without some evidence.
Vague bug reports are impossible to diagnose, but some general things apply.
- Yes, memory can be bad (one bad bit in one module is enough), but MUCH more common are software bugs. Moreover, changing memory configuration can provoke bugs that might otherwise lie hidden, because the OS and the apps behave differently with more (or less) memory.
- GPU drivers (by Apple) haven been notoriously buggy. Witness the 2013 Mac Pro debacle in which Adobe had no option but to remove GPU support for sharpening, after adding it.
- The software stack these days is extremely complex. Lightroom (or Photoshop) on top of GPU drivers in a multithreaded environment. It doesn’t take much to whack a pointer, the effects of which can show up seconds or minutes or hours later. Also, Adobe has had some reliability issues with Lightroom in the past year.
The fact so many problems occur and that the Mac also crashes with kernel panic (also seen) and with different memory (OWC vs Crucial) suggests a more general problem. There could just be something defective about this particular iMac.
Testing for bad memory
- diglloydTools MemoryTester is one option for checking core reliability (ideally, use the command-line version in Terminal, quit the Finder and all other apps so as to free up as much memory as possible).
- Run Apple Diagnostics. Run BOTH because sometimes problems are intermittent; diglloydTools MemoryTester can be left to run for a long period of time, sometimes detecting a problem.
- Swap out 3 of 4 memory modules; if the theory is a bad module, this can isolate the module. Tedious, but effective if one bad module. Or, start with 2 modules, then add one.
I don’t know for sure what is causing Hank M’s crashes. But I have some evidence-based guesses from the details he has provided. From what I see, there is a bug in code using a bad (invalid) pointer. That usually means that (a) code is overwriting memory, or (b) code is using pointers that have been deallocated to the free pool, or (c) some portion of the stack is overwritten, or similar. In other words, a plain 'ol software BUG. In other words, NOT bad memory and NOT a bad iMac. I’m speaking probabilities here; bad memory cannot be ruled out, but I see no evidence to support that.
For example, using Topaz InFocus (a sharpening tool) in Photoshop on a layer with a layername longer than 64 characters corrupts memory in Photoshop (Topaz refuses to fix this bug). Photoshop will crash 5-10-60 seconds later, depending. I’ve had to write a script to rename the layer to a short name, do the InFocus sharpening then rename the layer back to the original name!
Below, I don’t know what GetWarpedSourcePipe_A() isdoes but it looks to be in image processing code, and I know from personal experience that Adobe has had recent JPEG crash bugs inPhotoshop (fixed in latest release, at least the bug I encountered). So it would not surprise me that LR has crash bugs too, particularly with JPEG processing. But whether GetWarpedSourcePipe_A() is the culprit or it has been handed a bad pointer to work with cannot be aid from this stack crawl. Moreover, some bugs overwrite memory, and then the crash happens sometime later, when hapless code uses data that has been overwritten. So such things can be 'interesting' to track down.
Process: Adobe Lightroom  Path: /Applications/Adobe Lightroom/Adobe Lightroom.app/Contents/MacOS/Adobe Lightroom Identifier: com.adobe.Lightroom6 Version: Adobe Lightroom  (6.5) Code Type: X86-64 (Native) Parent Process: launchd  Responsible: Adobe Lightroom  User ID: 0 Date/Time: 2016-05-16 12:52:13.822 +0200 OS Version: Mac OS X 10.11.4 (15E65) Report Version: 11 Anonymous UUID: 44950BD0-1A4B-0CE2-ACCE-E955959A77A1 Time Awake Since Boot: 1800 seconds System Integrity Protection: disabled Crashed Thread: 43 GetWarpedSourcePipe_A (thread index 1) Exception Type: EXC_BAD_ACCESS (SIGSEGV) Exception Codes: KERN_INVALID_ADDRESS at 0x00007ff8d6b8bc10 Exception Note: EXC_CORPSE_NOTIFY VM Regions Near 0x7ff8d6b8bc10: MALLOC_TINY 00007fb8e2000000-00007fb8e3800000 [ 24.0M] rw-/rwx SM=PRV --> STACK GUARD 00007fff4c084000-00007fff4f884000 [ 56.0M] ---/rwx SM=NUL stack guard for thread 0 Thread 43 Crashed:: GetWarpedSourcePipe_A (thread index 1)
0 com.adobe.ag.AgImageIO 0x000000011736f464 0x116f90000 + 4060260
1 com.adobe.ag.AgImageIO 0x00000001170b5ac4 0x116f90000 + 1202884
2 com.adobe.ag.AgImageIO 0x00000001170a04b2 0x116f90000 + 1115314
3 com.adobe.ag.AgImageIO 0x00000001170b5e22 0x116f90000 + 1203746
4 com.adobe.ag.AgImageIO 0x00000001170a1635 0x116f90000 + 1119797
5 com.adobe.ag.AgImageIO 0x00000001171d3685 0x116f90000 + 2373253
6 com.adobe.ag.AgImageIO 0x00000001171d99a1 0x116f90000 + 2398625
7 com.adobe.ag.AgImageIO 0x00000001170b9915 0x116f90000 + 1218837
8 com.adobe.ag.AgImageIO 0x0000000117141ed2 0x116f90000 + 1777362
9 com.adobe.ag.AgImageIO 0x0000000117141de1 0x116f90000 + 1777121
10 libsystem_pthread.dylib 0x00007fff92b5699d _pthread_body + 131
11 libsystem_pthread.dylib 0x00007fff92b5691a _pthread_start + 168
12 libsystem_pthread.dylib 0x00007fff92b54351 thread_start + 13