April 2013 - Pro-apps-dev - lists.apple.com

Re: numerous kernel panics when using OpenCL inside FCPX
by Paul Miller 26 Apr '13

26 Apr '13

On 4/26/2013 1:47 PM, Garrick Meeker wrote: Some drivers do have bugs. The AMD driver in 10.6, in particular, may not be worth supporting. It doesn't support OpenCL images (requiring buffers only) and while we did work around that, we found the OpenCL/OpenGL interop so buggy that we disabled FxPlug hardware rendering for these cards/OS (instead transferring from host memory to the card). AMD 6xxx cards in 10.7 had issues (something I found with local memory) but an update seemed to fix that. However, you might be writing out of bounds to images or buffers, and very bad things can happen. Perhaps the first way to check this is to use the CPU device, call clEnqueueMapBuffer/clEnqueueMapImage on all buffers to print out the address, and see if you get a segfault (and compare that against your known buffer addresses). BTW, I'm about to open source a OpenCL logging utility that might interest you. It won't detect buffer overwrites (and kernel panics won't leave any log), but it's useful to track down errors on the host side like event handling. _______________________________________________ Do not post admin requests to the list. They will be ignored. Pro-apps-dev mailing list (Pro-apps-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: https://lists.apple.com/mailman/options/pro-apps-dev/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com I should have mentioned this is on an NV 330M. I'm getting a *crash* (instead of the full-on panic) in the same spot on an NV 8600M. I'm falling back to software if images aren't supported. I don't think this is happening, unless FCPX is lying about the bounds of the images. In this particular case, I'm getting a 656x487 FLOAT buffer from FCPX, and I'm creating an OCL image buffer of the same size and writing the image and then reading back to the output image of the same size. I am running a swizzle kernel, but that's it. The crash happens when I read back. Yeah, I could use something like that that. I've developed a bunch of my own logging/config stuff as well. The big trick is trying to make it all "just work", so we don't have to have a bunch of user-adjustable options.

1 0

Re: numerous kernel panics when using OpenCL inside FCPX
by Garrick Meeker 26 Apr '13

26 Apr '13

Some drivers do have bugs. The AMD driver in 10.6, in particular, may not be worth supporting. It doesn't support OpenCL images (requiring buffers only) and while we did work around that, we found the OpenCL/OpenGL interop so buggy that we disabled FxPlug hardware rendering for these cards/OS (instead transferring from host memory to the card). AMD 6xxx cards in 10.7 had issues (something I found with local memory) but an update seemed to fix that. However, you might be writing out of bounds to images or buffers, and very bad things can happen. Perhaps the first way to check this is to use the CPU device, call clEnqueueMapBuffer/clEnqueueMapImage on all buffers to print out the address, and see if you get a segfault (and compare that against your known buffer addresses). If that's not feasible, one other solution is to add manual bounds checking to all of your kernels. Consider this following code. You can add "-D BOUNDS_CHECK" to the compile options, add optional parameters to your host code (passing the additional int4 buffer and buffer sizes), add CHECK_ARG() and CHECK_ARRAY_ARG(a) to function arguments, and CHECK_ARRAY_DECL() for any private arrays. Then call CHECK_A() for array access and read_imagef_BOUNDS/write_imagef_BOUNDS for image access. Then you can turn this on and off (even at runtime). BTW, I'm about to open source a OpenCL logging utility that might interest you. It won't detect buffer overwrites (and kernel panics won't leave any log), but it's useful to track down errors on the host side like event handling. #ifndef STRCAT #define STRCAT(a,b) a ## b #endif #ifdef BOUNDS_CHECK int _CHECK_A(int i, int s, int line, __global int4 *bounds_log) { if ((i < 0) || i >= s) { bounds_log[0] = (int4)(1, i, 0, line); bounds_log[1] = (int4)(s, 0, 0, 0); i = max(0, min(s-1, i)); } return i; } int2 _CHECK_BOUNDS(image2d_t i, int2 s, int2 p, bool v, int line, __global int4 *bounds_log) { if (v && any((p < (int2)(0)) || p >= s)) { bounds_log[0] = (int4)(1, p.x, p.y, line); bounds_log[1] = (int4)(s.x, s.y, 0, 0); p = max((int2)(0), min(s-(int2)(1), p)); } return p; } #define CHECK_ARG() __global int4 *bounds_log, #define CHECK_ARRAY_DECL(a) const int STRCAT(a,Size) #define CHECK_ARRAY_ARG(a) a, int STRCAT(a,Size) #define CHECK_A(i, s) _CHECK_A((int)(i), (int)(s), __LINE__, bounds_log) #define CHECK_BOUNDS(i, p, v) _CHECK_BOUNDS(i, get_image_dim(i), p, v, __LINE__, bounds_log) #else #define CHECK_ARG() #define CHECK_ARRAY_ARG(a) a #define CHECK_A(i, s) (i) #define CHECK_BOUNDS(i, p, v) (p) #endif #ifndef CLK_ADDRESS_MASK #if defined(__CL_VERSION_1_1__) && (__OPENCL_VERSION__ >= __CL_VERSION_1_1__) #define CLK_ADDRESS_MASK (CLK_ADDRESS_NONE | CLK_ADDRESS_CLAMP | CLK_ADDRESS_CLAMP_TO_EDGE | CLK_ADDRESS_REPEAT | CLK_ADDRESS_MIRRORED_REPEAT) #else #define CLK_ADDRESS_MASK (CLK_ADDRESS_NONE | CLK_ADDRESS_CLAMP | CLK_ADDRESS_CLAMP_TO_EDGE | CLK_ADDRESS_REPEAT) #endif #endif #define read_imagef_BOUNDS(i,s,p) read_imagef(i,s,CHECK_BOUNDS(i,p,(s&CLK_ADDRESS_MASK)==CLK_ADDRESS_NONE)) #define write_imagef_BOUNDS(i,p,v) write_imagef(i,CHECK_BOUNDS(i,p,true),v) On Apr 25, 2013, at 2:34 PM, Paul Miller <stelefx(a)gmail.com> wrote: > Everything is working in my test code and in other hosts, but in FCPX when I attempt to run some OCL code on *D1* clips I get a repeatable kernel panic and have to restart the machine. > > Works fine on *HD* clips in FCPX. Works fine in Motion. > > Has anyone else run into this situation? > _______________________________________________ > Do not post admin requests to the list. They will be ignored. > Pro-apps-dev mailing list (Pro-apps-dev(a)lists.apple.com) > Help/Unsubscribe/Update your Subscription: > https://lists.apple.com/mailman/options/pro-apps-dev/garrick%40digitalanarc… > > This email sent to garrick(a)digitalanarchy.com _______________________________________________ Do not post admin requests to the list. They will be ignored. Pro-apps-dev mailing list (Pro-apps-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: https://lists.apple.com/mailman/options/pro-apps-dev/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com

1 0

numerous kernel panics when using OpenCL inside FCPX
by Paul Miller 25 Apr '13

25 Apr '13

Works fine on *HD* clips in FCPX. Works fine in Motion. Has anyone else run into this situation? _______________________________________________ Do not post admin requests to the list. They will be ignored. Pro-apps-dev mailing list (Pro-apps-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: https://lists.apple.com/mailman/options/pro-apps-dev/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com Everything is working in my test code and in other hosts, but in FCPX when I attempt to run some OCL code on *D1* clips I get a repeatable kernel panic and have to restart the machine.

1 0

Re: where to put license nags?
by Darrin Cardani 24 Apr '13

24 Apr '13

On Apr 24, 2013, at 1:29 PM, Paul Miller <stelefx(a)gmail.com> wrote: > We have been doing license checks and trial expiration nags in initWithApiManager(), but in FCPX this causes weirdness when dragging the effect, since it attempts to display a message box during the drag and usually aborts the drag. > > Darrin had suggested moving the license check to finishInitialSetup(), which works, but this is not called when effects are loaded from a saved setup. > > Has anyone solved this problem to their satisfaction? What about doing the check in initWithAPIManager, but not putting up the dialog until later? Or simply watermarking the output and having a "register" button (or whatever) that the user can click to bring up the registration dialog? Darrin _______________________________________________ Do not post admin requests to the list. They will be ignored. Pro-apps-dev mailing list (Pro-apps-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: https://lists.apple.com/mailman/options/pro-apps-dev/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com

1 0

where to put license nags?
by Paul Miller 24 Apr '13

24 Apr '13

Has anyone solved this problem to their satisfaction? _______________________________________________ Do not post admin requests to the list. They will be ignored. Pro-apps-dev mailing list (Pro-apps-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: https://lists.apple.com/mailman/options/pro-apps-dev/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com We have been doing license checks and trial expiration nags in initWithApiManager(), but in FCPX this causes weirdness when dragging the effect, since it attempts to display a message box during the drag and usually aborts the drag. Darrin had suggested moving the license check to finishInitialSetup(), which works, but this is not called when effects are loaded from a saved setup.

1 0

Re: one pixel shift to the right for CPU effects?
by Paul Miller 24 Apr '13

24 Apr '13

On 4/24/2013 1:05 PM, Darrin Cardani wrote: I've not seen this. If you can make a simple reproducible case and attach it to a bug, I'll look into it. _______________________________________________ Do not post admin requests to the list. They will be ignored. Pro-apps-dev mailing list (Pro-apps-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: https://lists.apple.com/mailman/options/pro-apps-dev/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com It only appears to happen on 1080P footage. I still haven't figured out why it only happens with some filters though.

1 0

Re: one pixel shift to the right for CPU effects?
by Darrin Cardani 24 Apr '13

24 Apr '13

Paul, I've not seen this. If you can make a simple reproducible case and attach it to a bug, I'll look into it. Darrin On Apr 24, 2013, at 5:25 AM, Paul Miller <paul(a)fxtech.com> wrote: > I'm noticing that some of my filters exhibit a single pixel shift to the right - even when parameters are set to "do nothing" values. But it's only on some of them. I can't find a rhyme or reason to the difference. > > Has anyone else seen anything like this? > > This is in 10.0.8. I didn't notice it in 10.0.7. > > _______________________________________________ > Do not post admin requests to the list. They will be ignored. > Pro-apps-dev mailing list (Pro-apps-dev(a)lists.apple.com) > Help/Unsubscribe/Update your Subscription: > https://lists.apple.com/mailman/options/pro-apps-dev/dcardani%40apple.com > > This email sent to dcardani(a)apple.com _______________________________________________ Do not post admin requests to the list. They will be ignored. Pro-apps-dev mailing list (Pro-apps-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: https://lists.apple.com/mailman/options/pro-apps-dev/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com

1 0

Re: Are new Motion templates working with older version of FCP X?
by Christoph Vonrhein 24 Apr '13

24 Apr '13

_______________________________________________ Do not post admin requests to the list. They will be ignored. Pro-apps-dev mailing list (Pro-apps-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: https://lists.apple.com/mailman/options/pro-apps-dev/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com

1 0

one pixel shift to the right for CPU effects?
by Paul Miller 24 Apr '13

24 Apr '13

Has anyone else seen anything like this? This is in 10.0.8. I didn't notice it in 10.0.7. _______________________________________________ Do not post admin requests to the list. They will be ignored. Pro-apps-dev mailing list (Pro-apps-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: https://lists.apple.com/mailman/options/pro-apps-dev/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com I'm noticing that some of my filters exhibit a single pixel shift to the right - even when parameters are set to "do nothing" values. But it's only on some of them. I can't find a rhyme or reason to the difference.

1 0

Re: Are new Motion templates working with older version of FCP X?
by Martin Baker 24 Apr '13

24 Apr '13

_______________________________________________ Do not post admin requests to the list. They will be ignored. Pro-apps-dev mailing list (Pro-apps-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: https://lists.apple.com/mailman/options/pro-apps-dev/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com

1 0