I mean:
After 1.7.30 engine of VST works fine, but GUI/editor just don't appear.

This shows in console when I'm trying open editor
fixme:dwrite:dwritetextlayout_GetMetrics (0x1b93a0)->(0x32efe4): stub

When you get a segfault you most likely also get a backtrace - please attach it to this bug report.

Moreover, please try to do a regression test for both commits. Would be useful to know which commit was responsible for breaking it, and which one partially restored it. Take a look here for more details:

http://wiki.winehq.org/RegressionTesting

I belive that Henri Verbeet, who fix previous problem can fix this even without regresion tests.

(In reply to Xj from comment #4)
> I belive that Henri Verbeet, who fix previous problem can fix this even
> without regresion tests.

It's very selfish to assume someone else can fix the issue without doing any work to help narrow it down for them.

Actually, it's not all that clear to me what this bug is about. The original crash in bug 36860 was fixed by commit c903192243e4438ab08e5426e24604b93c231493, after which the next bug was bug 36939. That one was fixed by commit 383854defae38574d69a56b5e361ba11c387ee63, after which as far as I can tell the plugin loads correctly, but doesn't draw correctly.

As mentioned in https://bugs.winehq.org/show_bug.cgi?id=36860#c4, that's pretty much expected since Direct2D is still far from done. The workaround of disabling d2d1 in winecfg should still work.

If this is purely about the drawing not working, that's fine, but not really a regression. We'd probably also want a d2d Bugzilla component.

(In reply to Henri Verbeet from comment #6)
> If this is purely about the drawing not working, that's fine, but not really
> a regression. We'd probably also want a d2d Bugzilla component.

Done.

Hello Henri. Yes the problem is drawing, actualy it doesn't draw GUI at all. And yes, the d2d workaround help, but you still want similar reports about this component, right ? I attach screenshot.

BTW. I see that there is more fixme messages ( sort|uniq version )
fixme:d2d:d2d_factory_CreateHwndRenderTarget iface 0x120190, desc 0x32f29c, hwnd_rt_desc 0x32f278, render_target 0x32f2bc stub!
fixme:d2d:d2d_factory_init Ignoring factory type and options.
fixme:d2d:d2d_stroke_style_init Ignoring stroke style properties.
fixme:dwrite:dwritetextlayout_GetMetrics (0x1baf70)->(0x32efe4): stub
fixme:dwrite:dwritetextlayout_GetMetrics (0x1fdac0)->(0x32efe4): stub
fixme:dwrite:get_name_record_codepage unknown platform 0
fixme:dwrite:get_name_record_locale unknown platform 0
fixme:ole:RemUnknown_QueryInterface No interface for iid {00000019-0000-0000-c000-000000000046}

BTW 2. Austin please stop judge peoples, cause you know nothing about this case, and you know nothing about me. It's very infantile. At least, I was the person who made previous regresion test. I'm quite sure that this time Henri know where is a problem, but if not I can try find wrong commit.

Created attachment 50011 [details]
Molot GUI crash

Created attachment 50564 [details]
vst gui bug - wine 1.7.35

I can confirm this bug in Wine 1.7.35. 
Other plugins affected too, for example Toneboosters Barricade v3.0.9 x86.
(http://www.toneboosters.com/download/)

I attached a screenshot.

Hello folks,

refining the summary, unfortunately being quite generic.
This should be avoided, making it a feature = meta bug.

Regards

Problem is still unsolved but new kind of debug appear - hopes this can help somehow:

fixme:d2d:d2d_factory_CreateHwndRenderTarget iface 0x148338, desc 0x44e10c, hwnd_rt_desc 0x44e0e8, render_target 0x44e12c stub!
fixme:d2d:d2d_stroke_style_init Ignoring stroke style properties.
fixme:d2d:d2d_factory_CreateHwndRenderTarget iface 0x148338, desc 0x44e10c, hwnd_rt_desc 0x44e0e8, render_target 0x44e12c stub!
fixme:d2d:d2d_stroke_style_init Ignoring stroke style properties.

(In reply to Xj from comment #13)
> Problem is still unsolved but new kind of debug appear - hopes this can help
> somehow:
> 
> fixme:d2d:d2d_factory_CreateHwndRenderTarget iface 0x148338, desc 0x44e10c,
> hwnd_rt_desc 0x44e0e8, render_target 0x44e12c stub!
> fixme:d2d:d2d_stroke_style_init Ignoring stroke style properties.
> fixme:d2d:d2d_factory_CreateHwndRenderTarget iface 0x148338, desc 0x44e10c,
> hwnd_rt_desc 0x44e0e8, render_target 0x44e12c stub!
> fixme:d2d:d2d_stroke_style_init Ignoring stroke style properties.

This looks to me bug 36499. I think it's quite safe to mark dupe.

Could someone link a simple freely available host application that I can use to load these plugins into?

(In reply to Nikolay Sivov from comment #15)
> Could someone link a simple freely available host application that I can use
> to load these plugins into?

Would LMMS work? 

https://lmms.io/download/

Commented too soon, it's easy to test with Reaper 0.999. Molot VST GUI renders, main problem is that some elements are missing. Problem seems to be related to d2d bitmap creation failure because underlying wic bitmap is too large (88 x 9064 as an example).

I'm closing this one, please open another report for rendering problem I mentioned.

*** This bug has been marked as a duplicate of bug 36499 ***

Original report was about that Molot doesn't work. First because of segfault and later because it can't draw GUI ( on default wine configuration ). Then there was a few bureaucratic changes like mark report as duplicate and finally it was mark as solved.

but .. nothing is solved - Molot will still doesnt' work for regular users without a hack (disable d2d1). No you say that it need another report for the same application, for the same problem ?

BTW. I used fsthost for host this plugin - download reaper just for test is overkill IMHO.

Did you try with current wine, 1.9.21? I works for me and doesn't crash when I try it with Reaper. Some UI issues I mentioned deserve a separate report, like I already said.

Not yet. I use wine-staging in Arch, but it's 1.9.20 currently. I'll give it a try ASAP, but I don't understand one thing. So if you know that there are issues with this application why it cannot be solved in this report ? Why we need to wait for another report ?

We try to have one report per Wine issue, not one per application that doesn't work, so we can track Wine progress looking at what was fixed. Also as it often happens applications suffer from several Wine bugs at once, one revealed after another one gets fixed.

Regarding this particular plugin I'm now getting texture creation failure in wined3d:

---
 trace:d3d11:d3d10_device_CreateTexture2D iface 0x172776c, desc 0x32a244, data 0x32a238, texture 0x32a234.
12556 trace:d3d:wined3d_texture_create device 0x280f640, desc 0x32a138, layer_count 1, level_count 1, flags 0, data 0x32a238, parent 0x2851320, parent_ops 0x7d074cac, texture 0x2851338.
12557 warn:d3d:texture_init Dimensions (88x9064) exceed the maximum texture size.
12558 warn:d3d:wined3d_texture_create Failed to initialize texture, returning 0x8876086a.
12559 warn:d3d11:d3d_texture2d_init Failed to create wined3d texture, hr 0x8876086a.
12560 warn:d3d11:d3d_texture2d_create Failed to initialize texture, hr 0x80070057.
12561 err:d2d:d2d_bitmap_create Failed to create texture, hr 0x80070057.
---

This is a correct PNG resource size, used by application to draw gauge indicator arrow for example. Original report was for a general d2d inability to create/use particular render target type.

Closing duplicate bugs.