Myujth

I don't think this is a problem with the image I'm loading. The image resolution is 256x256 so it's not the power of 2 issue. I've looked at other segfaults people got with glTexImage2D and the segfault still can't seem to go away. Sorry that the code isn't in C/C++ (i don't think this is the problem either) but it should still be easy to understand.

let surface = sdlimage.load("image.png") # equivalent to IMG_Load call in C
if surface.isNil:
 echo "Image couldn't be loaded: ", sdl2.getError()
 quit 1

glBlendFunc(GL_SRC_ALPHA, GL_ONE_MINUS_SRC_ALPHA)
glEnable(GL_BLEND)

var tex: cuint
glGenTextures(1, addr tex)
glBindTexture(GL_TEXTURE_2D, tex)
glPixelStorei(GL_UNPACK_ALIGNMENT, 1)
glPixelStorei(GL_UNPACK_ROW_LENGTH, surface.pitch)
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_LINEAR)
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, GL_LINEAR)
glTexImage2D(GL_TEXTURE_2D, 0, GL_RGBA.GLint, surface.w.GLsizei, surface.h.GLsizei, 0, GL_RGBA, GL_UNSIGNED_BYTE, surface.pixels) # segfault

UPDATE:

I managed to get it to not segfault thanks to user1118321's answer (I checked the pixel format and saw that SDL used a "indexed" format, which I saw another user said was the problem when they fixed this issue for theirselves. Seems like creating a new surface is the right solution), but now it shows nothing on the screen. It's black when I set glClearColor to black, and it's white when i set glClearColor to white.

Updated image loading code:

let surface = sdlimage.load("image.png")
if surface.isNil:
 echo "Image couldn't be loaded: ", sdl2.getError()
 quit 1

var w = surface.w # may need to make this the next power of two
var h = surface.h # and this
var bpp: cint
var Rmask, Gmask, Bmask, Amask: uint32
if not pixelFormatEnumToMasks(SDL_PIXELFORMAT_ABGR8888, bpp,
 Rmask, Gmask, Bmask, Amask):
 quit "pixel format enum to masks " & $sdl2.getError()

let newSurface = createRGBSurface(0, w, h, bpp,
 Rmask, Gmask, Bmask, Amask)

discard surface.setSurfaceAlphaMod(0xFF)
discard surface.setSurfaceBlendMode(BlendMode_None)

blitSurface(surface, nil, newSurface, nil)

var tex: cuint
glGenTextures(1, addr tex)
glBindTexture(GL_TEXTURE_2D, tex)
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_LINEAR)
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, GL_LINEAR)
glTexImage2D(GL_TEXTURE_2D, 0, 4, w.GLsizei, h.GLsizei, 0, GL_RGBA, GL_UNSIGNED_BYTE, newSurface.pixels)

Image rendering code:

glUseProgram(shaderProgram)
glClear(GL_COLOR_BUFFER_BIT or GL_DEPTH_BUFFER_BIT)
glMatrixMode(GL_MODELVIEW)
glLoadIdentity()
glTranslatef(0.0, 0.0, -7.0)
glColor3f(1.0, 1.0, 1.0)
glBegin(GL_QUADS)
glTexCoord2i(0, 0)
glVertex3f(-0.5, -1.9, 0.0)
glTexCoord2i(1, 0)
glVertex3f(0.5, -1.9, 0.0)
glTexCoord2i(0, 1)
glVertex3f(-0.5, -1.4, 0.0)
glTexCoord2i(1, 1)
glVertex3f(0.5, -1.4, 0.0)
glEnd()

If shaders have something to do with it, here's my vertex shader:

#version 330 core
in vec4 data;
out vec3 ourColor;
void main() 
 gl_Position = vec4(data.x, data.y, data.z, 1.0);
 ourColor = vec3(data.w, data.w, data.w);

And the fragment shader:

#version 330 core
precision highp float;
in vec3 ourColor;
out vec4 color;
void main() 
 color = vec4(ourColor, 1.0);

You should check glGetError() to see if there are any OpenGL errors occurring. You should also check the surface.pixelFormat to see whether you actually have 4 bytes per pixel in your image. If it's just RGB data and not RGBA, then glTexImage2D() will read past the end of the image data and likely crash with a segmentation fault.