glReadPixels from a pixel buffer object yields an empty buffer

[cratonica]

Is this the right place for my bug report?
Calls are done through the userland GL libraries, so I'm starting here. It could be a firmware bug though.

Describe the bug
Rendering to pixel buffer objects (or potentially readback from them?) don't seem to work as it does on other devices. Instead, it yields a blank buffer when it should contain the framebuffer that was rendered. On other Linux devices, this works as intended. It also works when rendering directly to the framebuffer and then calling glReadPixels, but I'd like to avoid stalling the graphics pipeline as recommended here: http://www.songho.ca/opengl/gl_pbo.html.

The example program from the link above exhibits this same issue, but I've attached a simpler example to reproduce the problem below.

To reproduce
Compile the following code twice, once with pixel buffer rendering and once without:

c++ -std=c++17 gltest.cpp -lglut -lGL -o standard
c++ -std=c++17 gltest.cpp -lglut -lGL -o pbo -DPBO

#include <cstdint>
#include <cstdio>
#include <time.h>
#include <cassert>

#ifdef PBO
#define GL_GLEXT_PROTOTYPES
#include <GL/gl.h>
#include <GL/glext.h>
#endif
#include <GL/glut.h>

constexpr int WIDTH = 256;
constexpr int HEIGHT = 256;
constexpr double ASPECT_RATIO = double(WIDTH) / double(HEIGHT);
constexpr int FRAMEBUFFER_BYTES = WIDTH * HEIGHT * 4;

#ifdef PBO
static GLuint pboIds[2];
static int index;
static int nextIndex;
#endif

void DrawMesh() {
    glBegin(GL_QUADS);
        glColor4f(1, 0, 0, 1);
        glVertex3f(-1.0, -1.0, 0.0);
        glColor4f(0, 1, 0, 1);
        glVertex3f(1, -1, 0.0);
        glColor4f(0, 0, 1, 1);
        glVertex3f(1, 1, 0.0);
        glColor4f(1, 0, 1, 1);
        glVertex3f(-1, 1, 0.0);
    glEnd();
}

float rotation = 0.0f;
bool auto_rotate = true;

#ifndef PBO
uint8_t framebuffer[FRAMEBUFFER_BYTES];
#endif

void Render() {
    glClearColor(0, 0, 0, 1);
    glClear(GL_COLOR_BUFFER_BIT | GL_DEPTH_BUFFER_BIT);

    glMatrixMode(GL_PROJECTION);
    glLoadIdentity();
    glOrtho(-1.5 * ASPECT_RATIO, 1.5 * ASPECT_RATIO, -1.5, 1.5, -1.5, 20);

    glMatrixMode(GL_MODELVIEW);
    glLoadIdentity();
    glRotatef(rotation, 0, 0, 1);
    DrawMesh();
    glutSwapBuffers();
}

void Idle() {
    if (auto_rotate) {
        rotation += 0.5f;
    }
    glutPostRedisplay();

#ifdef PBO
    index = (index + 1) % 2;
    nextIndex = (index + 1) % 2;
#endif

    struct timespec start, end;
    clock_gettime(CLOCK_MONOTONIC, &start);

#ifdef PBO
    glReadBuffer(GL_FRONT);
    glBindBuffer(GL_PIXEL_PACK_BUFFER, pboIds[index]);
    glReadPixels(0, 0, WIDTH, HEIGHT, GL_RGBA, GL_UNSIGNED_BYTE, 0);
    glBindBuffer(GL_PIXEL_PACK_BUFFER, pboIds[nextIndex]);
    GLubyte* framebuffer = (GLubyte*)glMapBuffer(GL_PIXEL_PACK_BUFFER, GL_READ_ONLY);
#else
    glReadPixels(0, 0, WIDTH, HEIGHT, GL_RGBA, GL_UNSIGNED_INT_8_8_8_8_REV, framebuffer);
#endif

    clock_gettime(CLOCK_MONOTONIC, &end);
    const long long nanos =
      1000000000LL * (end.tv_sec - start.tv_sec) + end.tv_nsec - start.tv_nsec;
    printf("glReadPixels in %Ldus\n", nanos / (1000LL));

    static int cnt;
    ++cnt;
    char fname[128];
    sprintf(fname, "/tmp/frame%d.data", cnt);
    FILE* fout = fopen(fname, "wb");
    fwrite(framebuffer, 1, FRAMEBUFFER_BYTES, fout);
    fclose(fout);

#ifdef PBO
  glUnmapBuffer(GL_PIXEL_PACK_BUFFER);
  glBindBuffer(GL_PIXEL_PACK_BUFFER, 0);
#endif
}

int main(int argc, char** argv) {
    glutInit(&argc, argv);
    glutInitDisplayMode(GLUT_DOUBLE);
    glutInitWindowSize(WIDTH, HEIGHT);
    glutInitWindowPosition(256, 256);
    glutCreateWindow("GL Test");
    glutDisplayFunc(Render);
    glutIdleFunc(Idle);

#ifdef PBO
    glGenBuffers(2, pboIds);
    glBindBuffer(GL_PIXEL_UNPACK_BUFFER, pboIds[0]);
    glBufferData(GL_PIXEL_UNPACK_BUFFER, FRAMEBUFFER_BYTES, 0, GL_STREAM_DRAW);
    glBindBuffer(GL_PIXEL_UNPACK_BUFFER, pboIds[1]);
    glBufferData(GL_PIXEL_UNPACK_BUFFER, FRAMEBUFFER_BYTES, 0, GL_STREAM_DRAW);
    glBindBuffer(GL_PIXEL_UNPACK_BUFFER, 0);
#endif

    glutMainLoop();
    return 0;
}

Expected behaviour
Both program variants should write 256x256 RGBA-format byte buffers /tmp/frameXX.data containing the framebuffer data of a multicolor square rotating.

Actual behaviour
Only the standard implementation writes the framebuffer correctly. The one that uses pixel buffer objects writes a blank framebuffer. You can md5sum /tmp/frame*.data and they will all have the same hash when running in PBO mode. You can view these files in GIMP by opening them as raw data, RGBA 256x256.

As mentioned, every other system I've run this code on works as expected.

System
Copy and paste the results of the raspinfo command in to this section.

System Information
------------------

Raspberry Pi Zero 2 W Rev 1.0
PRETTY_NAME="Raspbian GNU/Linux 11 (bullseye)"
NAME="Raspbian GNU/Linux"
VERSION_ID="11"
VERSION="11 (bullseye)"

Raspberry Pi reference 2022-01-28
Generated using pi-gen, https://github.com/RPi-Distro/pi-gen, f01430c9d8f67a4b9719cc00e74a2079d3834d5d, stage4

Linux raspberrypi 5.10.103-v7+ #1530 SMP Tue Mar 8 13:02:44 GMT 2022 armv7l GNU/Linux
Revision	: 902120
Serial		: 00000000d7638eed
Model		: Raspberry Pi Zero 2 W Rev 1.0
Throttled flag  : throttled=0x0
Camera          : supported=1 detected=0

Videocore information
---------------------

Jan 20 2022 13:59:16 
Copyright (c) 2012 Broadcom
version bd88f66f8952d34e4e0613a85c7a6d3da49e13e2 (clean) (release) (start_x)

alloc failures:     0
compactions:        0
legacy block fails: 0

Filesystem information
----------------------
Filesystem     1K-blocks    Used Available Use% Mounted on
/dev/root       29734180 4107792  24363088  15% /
devtmpfs          153364       0    153364   0% /dev
tmpfs             186644       0    186644   0% /dev/shm
tmpfs              74660     692     73968   1% /run
tmpfs               5120       4      5116   1% /run/lock
/dev/mmcblk0p1    258095   49440    208655  20% /boot
tmpfs              37328      20     37308   1% /run/user/1000

Filename				Type		Size		Used		Priority
/var/swap                               file		102396		49408		-2

Package version information
---------------------------
raspberrypi-ui-mods:
  Installed: 1.20220301
raspberrypi-sys-mods:
  Installed: 20220224
openbox:
  Installed: 3.6.1-9+rpt1+deb11u1
lxpanel:
  Installed: 0.10.1-2+rpt12
pcmanfm:
  Installed: 1.3.2-1+rpt6
rpd-plym-splash:
  Installed: 0.30

Output of glGetString(GL_EXTENSIONS) indicates that this should be supported (see GL_ARB_pixel_buffer_object below):

GL_ARB_multisample GL_EXT_abgr GL_EXT_bgra GL_EXT_blend_color GL_EXT_blend_minmax GL_EXT_blend_subtract GL_EXT_copy_texture GL_EXT_subtexture GL_EXT_texture_object GL_EXT_vertex_array GL_EXT_compiled_vertex_array GL_EXT_texture GL_EXT_texture3D GL_IBM_rasterpos_clip GL_ARB_point_parameters GL_EXT_draw_range_elements GL_EXT_packed_pixels GL_EXT_point_parameters GL_EXT_rescale_normal GL_EXT_separate_specular_color GL_EXT_texture_edge_clamp GL_SGIS_generate_mipmap GL_SGIS_texture_border_clamp GL_SGIS_texture_edge_clamp GL_SGIS_texture_lod GL_ARB_framebuffer_sRGB GL_ARB_multitexture GL_EXT_framebuffer_sRGB GL_IBM_multimode_draw_arrays GL_IBM_texture_mirrored_repeat GL_ARB_texture_cube_map GL_ARB_texture_env_add GL_ARB_transpose_matrix GL_EXT_blend_func_separate GL_EXT_fog_coord GL_EXT_multi_draw_arrays GL_EXT_secondary_color GL_EXT_texture_env_add GL_EXT_texture_lod_bias GL_INGR_blend_func_separate GL_NV_blend_square GL_NV_light_max_exponent GL_NV_texgen_reflection GL_NV_texture_env_combine4 GL_S3_s3tc GL_SUN_multi_draw_arrays GL_ARB_texture_border_clamp GL_ARB_texture_compression GL_EXT_framebuffer_object GL_EXT_texture_compression_s3tc GL_EXT_texture_env_combine GL_EXT_texture_env_dot3 GL_MESA_window_pos GL_NV_packed_depth_stencil GL_NV_texture_rectangle GL_ARB_depth_texture GL_ARB_occlusion_query GL_ARB_shadow GL_ARB_texture_env_combine GL_ARB_texture_env_crossbar GL_ARB_texture_env_dot3 GL_ARB_texture_mirrored_repeat GL_ARB_window_pos GL_ATI_fragment_shader GL_EXT_stencil_two_side GL_EXT_texture_cube_map GL_NV_copy_depth_to_color GL_NV_depth_clamp GL_NV_fog_distance GL_NV_half_float GL_APPLE_packed_pixels GL_ARB_draw_buffers GL_ARB_fragment_program GL_ARB_fragment_shader GL_ARB_shader_objects GL_ARB_vertex_program GL_ARB_vertex_shader GL_ATI_draw_buffers GL_ATI_texture_env_combine3 GL_ATI_texture_float GL_EXT_shadow_funcs GL_EXT_stencil_wrap GL_MESA_pack_invert GL_MESA_ycbcr_texture GL_NV_primitive_restart GL_ARB_depth_clamp GL_ARB_fragment_program_shadow GL_ARB_half_float_pixel GL_ARB_occlusion_query2 GL_ARB_point_sprite GL_ARB_shading_language_100 GL_ARB_sync GL_ARB_texture_non_power_of_two GL_ARB_vertex_buffer_object GL_ATI_blend_equation_separate GL_EXT_blend_equation_separate GL_OES_read_format GL_ARB_color_buffer_float GL_ARB_pixel_buffer_object GL_ARB_texture_compression_rgtc GL_ARB_texture_float GL_ARB_texture_rectangle GL_ATI_texture_compression_3dc GL_EXT_packed_float GL_EXT_pixel_buffer_object GL_EXT_texture_compression_dxt1 GL_EXT_texture_compression_rgtc GL_EXT_texture_mirror_clamp GL_EXT_texture_rectangle GL_EXT_texture_sRGB GL_EXT_texture_shared_exponent GL_ARB_framebuffer_object GL_EXT_framebuffer_blit GL_EXT_framebuffer_multisample GL_EXT_packed_depth_stencil GL_ARB_vertex_array_object GL_ATI_separate_stencil GL_ATI_texture_mirror_once GL_EXT_draw_buffers2 GL_EXT_draw_instanced GL_EXT_gpu_program_parameters GL_EXT_gpu_shader4 GL_EXT_texture_array GL_EXT_texture_compression_latc GL_EXT_texture_integer GL_EXT_texture_sRGB_decode GL_EXT_timer_query GL_OES_EGL_image GL_EXT_texture_buffer_object GL_AMD_texture_texture4 GL_ARB_copy_buffer GL_ARB_depth_buffer_float GL_ARB_draw_instanced GL_ARB_half_float_vertex GL_ARB_instanced_arrays GL_ARB_map_buffer_range GL_ARB_texture_buffer_object GL_ARB_texture_rg GL_ARB_texture_swizzle GL_ARB_vertex_array_bgra GL_EXT_texture_swizzle GL_EXT_vertex_array_bgra GL_NV_conditional_render GL_AMD_conservative_depth GL_AMD_draw_buffers_blend GL_AMD_seamless_cubemap_per_texture GL_AMD_shader_stencil_export GL_ARB_ES2_compatibility GL_ARB_blend_func_extended GL_ARB_compatibility GL_ARB_debug_output GL_ARB_draw_buffers_blend GL_ARB_draw_elements_base_vertex GL_ARB_explicit_attrib_location GL_ARB_fragment_coord_conventions GL_ARB_provoking_vertex GL_ARB_sample_shading GL_ARB_sampler_objects GL_ARB_seamless_cube_map GL_ARB_shader_stencil_export GL_ARB_shader_texture_lod GL_ARB_texture_buffer_object_rgb32 GL_ARB_texture_cube_map_array GL_ARB_texture_gather GL_ARB_texture_multisample GL_ARB_texture_query_lod GL_ARB_texture_rgb10_a2ui GL_ARB_uniform_buffer_object GL_ARB_vertex_type_2_10_10_10_rev GL_EXT_provoking_vertex GL_EXT_texture_snorm GL_MESA_texture_signed_rgba GL_NV_copy_image GL_NV_texture_barrier GL_ARB_draw_indirect GL_ARB_get_program_binary GL_ARB_robustness GL_ARB_separate_shader_objects GL_ARB_shader_bit_encoding GL_ARB_shader_subroutine GL_ARB_texture_compression_bptc GL_ARB_timer_query GL_ARB_transform_feedback2 GL_ARB_transform_feedback3 GL_ARB_viewport_array GL_EXT_direct_state_access GL_AMD_multi_draw_indirect GL_ANGLE_texture_compression_dxt3 GL_ANGLE_texture_compression_dxt5 GL_ARB_base_instance GL_ARB_compressed_texture_pixel_storage GL_ARB_conservative_depth GL_ARB_internalformat_query GL_ARB_map_buffer_alignment GL_ARB_shader_atomic_counters GL_ARB_shader_image_load_store GL_ARB_shading_language_420pack GL_ARB_shading_language_packing GL_ARB_texture_storage GL_ARB_transform_feedback_instanced GL_EXT_framebuffer_multisample_blit_scaled GL_EXT_transform_feedback GL_AMD_query_buffer_object GL_AMD_shader_trinary_minmax GL_ARB_ES3_compatibility GL_ARB_arrays_of_arrays GL_ARB_clear_buffer_object GL_ARB_compute_shader GL_ARB_copy_image GL_ARB_explicit_uniform_location GL_ARB_fragment_layer_viewport GL_ARB_framebuffer_no_attachments GL_ARB_invalidate_subdata GL_ARB_multi_draw_indirect GL_ARB_program_interface_query GL_ARB_robust_buffer_access_behavior GL_ARB_shader_image_size GL_ARB_shader_storage_buffer_object GL_ARB_stencil_texturing GL_ARB_texture_buffer_range GL_ARB_texture_query_levels GL_ARB_texture_storage_multisample GL_ARB_texture_view GL_ARB_vertex_attrib_binding GL_KHR_debug GL_KHR_robustness GL_KHR_texture_compression_astc_ldr GL_ARB_buffer_storage GL_ARB_clear_texture GL_ARB_enhanced_layouts GL_ARB_indirect_parameters GL_ARB_internalformat_query2 GL_ARB_multi_bind GL_ARB_query_buffer_object GL_ARB_seamless_cubemap_per_texture GL_ARB_shader_draw_parameters GL_ARB_shader_group_vote GL_ARB_shading_language_include GL_ARB_texture_mirror_clamp_to_edge GL_ARB_texture_stencil8 GL_ARB_vertex_type_10f_11f_11f_rev GL_EXT_shader_integer_mix GL_ARB_clip_control GL_ARB_conditional_render_inverted GL_ARB_cull_distance GL_ARB_derivative_control GL_ARB_direct_state_access GL_ARB_get_texture_sub_image GL_ARB_pipeline_statistics_query GL_ARB_shader_texture_image_samples GL_ARB_texture_barrier GL_ARB_transform_feedback_overflow_query GL_EXT_polygon_offset_clamp GL_KHR_blend_equation_advanced GL_KHR_context_flush_control GL_KHR_robust_buffer_access_behavior GL_ARB_parallel_shader_compile GL_ARB_post_depth_coverage GL_ARB_shader_atomic_counter_ops GL_KHR_no_error GL_KHR_texture_compression_astc_sliced_3d GL_MESA_shader_integer_functions GL_ARB_polygon_offset_clamp GL_KHR_parallel_shader_compile GL_EXT_EGL_image_storage GL_EXT_shader_framebuffer_fetch_non_coherent GL_EXT_texture_sRGB_R8 GL_EXT_texture_shadow_lod GL_EXT_EGL_sync