Add git_branch to git-version.h, add display version info to gs_frame and add branch and version to log. #3186

PeterMcteague · 2017-08-09T22:48:48Z

This was requested in this issue: #3152.

In this pull are the modifications required for it to appear.

Asinin3 · 2017-08-10T08:45:29Z

Thank you for this, it will indeed solve a lot of back and forth between users, I'm not sure if the whole version number with the commit# is necessary (at least for the master) though. I think just the build/commit number would be enough if the title is too cluttered (as some of the devs seemed to think).

hcorion · 2017-08-13T21:15:15Z

This is pretty ugly, and has too much information.
I'd like for it to be just commithash-branchname
So it would then be

784a37fc-vertex_rewrite

or if it was the master branch, then it would be just

784a37fc

Asinin3 · 2017-08-13T21:52:27Z

@hcorion Would it be possible to use the appveyor build numbers? Probably not because the number is only generated by the appveyor count right? But I think that would be the easiest to tell what build it is or give an estimation of how old a build is compared to latest while keeping it short.

For instance:
RPCS3 0.0.3-buildnumber-branchname
RPCS3 0.0.3-5680-vertex_rewrite

hcorion · 2017-08-13T23:19:40Z

Sure, I can agree with that, but without the RPCS3 part, so just 0.0.3-5680-vertex_rewrite

PeterMcteague · 2017-08-13T23:34:14Z

Alright, I've managed to change it to be in the format of 0.0.3-buildnumber however I'm struggling to get the branch name.

I assumed git-version.cmake and git-version.h would be the places to get it, after adding a command for getting the branch to git-version.cmake. However it seems whatever I change in the cmake file it has no effect on git-version.h.

I'll try and figure it out but if anyone's got any pointers they'd be appreciated.

hcorion · 2017-08-13T23:40:13Z

probably because normal Visual Studio builds don't regenerate git-version.h?
unless you're using linux.

Perhaps this git rev-parse --abbrev-ref HEAD could be helpful

PeterMcteague · 2017-08-15T20:19:21Z

I believe my latest commit should have this request completed.

I have updated the batch file to generate git-version numbers on windows with support for the branch number and have gotten the expected result on Windows.

Below is a picture of a game being ran on RPCS3 using this version. Note the version number at the top along with the branch name. Note the branch name intentionally should not appear if branch = "master" as suggested.

I have not tested on Linux but I believe the C++ code should work the same and the git-version.cmake file has been updated to also generate branch names in git-version.h , so I would expect it to work.

If needs be I can recreate the pull request for cleaner commits, as mine have been a little messy (Still getting used to things).

Asinin3 · 2017-08-16T20:55:12Z

Would it be possible for you to add the branchname to the RPCS3.log file? would help identify the build people are using there as well without much effort.

PeterMcteague · 2017-08-16T21:39:05Z

Yeah it should be easy enough I think. Might be a bit annoying to seperate my branch name stuff from the gs_frame stuff to commit it seperately, but I'll try.

Asinin3 · 2017-08-16T21:41:13Z

Well, you don't have to do it in this PR if it's messy, I just figured now was a good time to ask since you have experience with it now and hcorion seemed to be interested enough that he told me to ask you.

PeterMcteague · 2017-08-16T21:48:33Z

Cleaner if I do it in this PR, harder if I do it seperately. Will just do it in this PR.

EDIT: Done. Log file now looks like (Example from my log)

RPCS3 v0.0.3-5672-19612d21 Alpha
Branch: branch-in-logs
Intel(R) Core(TM) i5-4670K CPU @ 3.40GHz | 4 Threads | 15.92 GiB RAM | AVX

hcorion · 2017-08-18T19:58:54Z

Yep, works on Linux

snippet from log:

RPCS3 v0.0.3-5711-b256e281 Alpha
Branch: super-branch
Intel(R) Core(TM) i7-4770 CPU @ 3.40GHz | 8 Threads | 15.63 GiB RAM | AVX
·! 0:00:00.000000 LDR: Path: /home/zion/.config/rpcs3/dev_hdd0/game/NPUB90104/USRDIR/EBOOT.BIN
·! 0:00:00.000416 LDR: Title: 1942: Joint Strike
·! 0:00:00.000423 LDR: Serial: NPUB90104
·! 0:00:00.000446 LDR: Cache: /home/zion/.config/rpcs3/data/NPUB90104/75f7ec3436dd3a039b41c21d-EBOOT.BIN/
·! 0:00:00.000453 LDR: Applying custom config: data/NPUB90104/config.yml
·! 0:00:00.001516 LDR: Used configuration:

Megamouse · 2017-08-20T15:09:56Z

rpcs3/rpcs3qt/gs_frame.cpp

 	if (!Emu.GetTitle().empty())
 	{
-		m_windowTitle += qstr(" | " + Emu.GetTitle());


just wondering why you removed this and put it up there.
in case the ifdef fails we will have something like "VulkanDemon's Souls"

Ah the removal of the '|' is definitely a mistake, I don't actually recall why I did that. I'll fix it in a second.

hcorion · 2017-08-26T00:09:43Z

With this PR it now crashes RPCS3, before the GUI even launches, on Linux.

0x0000000000db3134 in logs::file_writer::log (this=0x4099300 <logs::get_logger()::logger>, size=0x3, text=<optimized out>) at /mnt/hd/Users/Zion/Documents/source/rpcs3-zion/Utilities/Log.cpp:327
327                     std::memcpy(m_fptr + pos, text, size);
gdb$ bt
#0  0x0000000000db3134 in logs::file_writer::log (this=0x4099300 <logs::get_logger()::logger>, size=0x3, text=<optimized out>) at /mnt/hd/Users/Zion/Documents/source/rpcs3-zion/Utilities/Log.cpp:327
#1  logs::file_listener::file_listener (this=0x4099300 <logs::get_logger()::logger>, name=...) at /mnt/hd/Users/Zion/Documents/source/rpcs3-zion/Utilities/Log.cpp:336
#2  0x0000000000db2669 in logs::get_logger () at /mnt/hd/Users/Zion/Documents/source/rpcs3-zion/Utilities/Log.cpp:85
#3  0x0000000000db2510 in logs::listener::add (_new=0x4093580 <s_gui_listener>) at /mnt/hd/Users/Zion/Documents/source/rpcs3-zion/Utilities/Log.cpp:203
#4  0x00000000007aa969 in _GLOBAL__sub_I_log_frame.cpp () at /mnt/hd/Users/Zion/Documents/source/rpcs3-zion/rpcs3/rpcs3qt/log_frame.cpp:42
#5  0x0000000001b7dc0d in __libc_csu_init ()
#6  0x00007fdb7dae845b in __libc_start_main () from /usr/lib/libc.so.6
#7  0x00000000007ff16a in _start () at /usr/include/qt/QtCore/qarraydata.h:237

hcorion · 2017-08-26T00:13:25Z

Utilities/Log.cpp

@@ -293,7 +294,6 @@ logs::file_writer::file_writer(const std::string& name)
 		m_fmap = CreateFileMappingW(m_file.get_handle(), 0, PAGE_READWRITE, s_log_size >> 32, s_log_size & 0xffffffff, 0);
 		m_fptr = (uchar*)MapViewOfFile(m_fmap, FILE_MAP_WRITE, 0, 0, 0);
 #else
-		m_file.trunc(s_log_size);


Don't delete this line, without this line RPCS3 crashes on Linux.

Ah must've deleted that by mistake when I was resolving the merge conflict. I've added it back in and will double check I haven't accidentally deleted anything else.

hcorion · 2017-08-28T17:08:47Z

I was talking to Nekotekina, and he said he'd like all the unnecessary changes to Log.cpp to be undone.

PeterMcteague · 2017-08-28T17:28:07Z

I've looked at the file changes and there's the addition of the branch in the log files and some spacing changes before ='s - which weren't intentionally done by me, it just seems to be how my visual studio has interpreted the file for whatever reason , as well as some new lines being moved because lines have been added.

Are the unintentional spacing changes what should be undone?

hcorion · 2017-08-28T17:41:52Z

I would assume so, yes.

PeterMcteague · 2017-08-30T17:16:31Z

The spacing changes are reverted in commit 4d22317.

scribam · 2017-09-09T19:48:01Z

Do you think you can squash your commits and rebase your branch onto master? It would be probably cleaner to merge.

PeterMcteague · 2017-09-09T19:55:35Z

Yes, I should be able to later if it makes things easier.

EDIT: Done! I believe them to be correct , I've got to test a second though to make sure I rebased without issues.

EDIT2: Yep , working as expected!

…lso involved making the files that generate git-version.h , get the branch.

hcorion

LGTM 👍

Nekotekina · 2017-09-13T18:19:38Z

Utilities/Log.cpp

-	ver.text  = fmt::format("RPCS3 v%s\n%s", rpcs3::version.to_string(), utils::get_system_info());
+
+	#ifdef RPCS3_GIT_BRANCH
+		ver.text = fmt::format("RPCS3 v%s\nBranch: %s\n%s\n", rpcs3::version.to_string(), RPCS3_GIT_BRANCH, utils::get_system_info());


I think it can be solved without copying this line.
Maybe add this to rpcs3::version then? Plus don't really need another line for this.

Ah yes , I seem to have missed that. I'll shortly change that to one line and put the branch into RPCS3::version.

EDIT: Having tried it , it seems when I move branch to rpcs3::version it doesn't get defined until after the log is created. I will commit cleaning up the lines and do some reading into that.

There we go. Thanks for the review , I think it should be better now! :)

These aren't neccesary as git_branch will definitely be defined, as it's generated.

-Version for gs_frame is now one line -Unncesary ifdefs removed again

Nekotekina

Bugs

Nekotekina · 2017-09-15T10:58:21Z

rpcs3/rpcs3_version.cpp

 namespace rpcs3
 {
 	const extern utils::version version{ 0, 0, 3, utils::version_type::alpha, 1, RPCS3_GIT_VERSION };
+	const char* branch{get_branch()};


This will be nullptr at startup and may crash depending on initialization order. It must be a function callable from the outside.

Think it should avoid that in the latest commit, not 100% sure though.

Nekotekina · 2017-09-15T10:58:59Z

rpcs3/rpcs3qt/gs_frame.cpp

+	version = version.substr(0 , version.find_last_of("-"));
+
+	//Add branch to version on frame , unless it's master.
+	if (rpcs3::branch != "master")


Not a string comparison

It's now inside the namespace and is just a function returning a std::string which presents less issues than the previous implementation.

Nekotekina · 2017-09-16T12:06:00Z

rpcs3/rpcs3_version.h

 	extern const utils::version version;
-}
+}


Don't remove newlines at the end of file

Alright , will keep an eye out for that in future.

Nekotekina · 2017-09-16T18:01:25Z

Utilities/Log.cpp

@@ -1,4 +1,4 @@
-#include "Log.h"
+#include "Log.h"


How you managed to damage unicode mark?

Couldn't tell you , I'm just using Visual Studio community 2017 on Windows. Would you like me to do another commit to try and remove these kind of commits (Removed X and added X) and remove some unintended changes like newlines being removed?

No, I pushed fixes in my PR

… and add branch and version to log. (RPCS3#3186) * Added version number and branch name to gs_frame and log file. This also involved making the files that generate git-version.h , get the branch.

* Qt: Remove duplicate gui repaints in settings dialog ApplyGuiOptions(true) already forces GuiStylesheetRequest and therefore RepaintGui in main_window if necessary. We don't need GuiRepaintRequest on accept because there should be no option left that doesn't repaint immediately * Qt: improve gamelist refresh, fix some more scroll issues despite earlier efforts with modelindex and so on the annoying scroll jump still surfaced sometimes, so we only scrollTo index on selected occasions from now on. Furthermore we don't need to refresh the gamelist if no custom config has been applied, so this will only be needed on accept from now on * Fix buffer overrun in cert loading, made it more accurate. Fixes RPCS3#3408 * Added additional disc folder detection and INSDIR functionality. * fs: minor fix * sys_fs: use g_tls_error for diagnostic Affected syscalls: sys_fs_open sys_fs_opendir sys_fs_stat sys_fs_mkdir sys_fs_fcntl (get free space) * sys_mmapper: minor range fix * Improve lv2_memory object * Implement fs::error::notempty * Vulkan: use pkg-config instead of checking for Linux (RPCS3#3440) Makes Vulkan optional on Linux / makes it possible on FreeBSD. * Update Visual Studio 2015 URL * Add git_branch to git-version.h, add display version info to gs_frame and add branch and version to log. (RPCS3#3186) * Added version number and branch name to gs_frame and log file. This also involved making the files that generate git-version.h , get the branch. * Implemented installation of PS3_EXTRA PS3_EXTRA games are installed if they're not already and then the game boots. Unsure if the game booting is the correct behavior but I think it is. * Redid a forgotten change * Qt: settings_dialog layout fix (hack) * Qt: add char count label to osk dialog * [Qt] fix graphics adapter selection and rebase for less redundancy * HGA calls sys_rsx_context_attribute before calling sys_rsx_memory_allocate. This inevitably causes RPCS3 to crash. Add nullptr check to prevent this. * Moved all currently written installs to main_window This gives a much nicer install dialog. In future I want to give some better way of installing the PS3_EXTRA stuff, preferably giving it it's own menu. * Moved installs to main_window so they could have a progress bar This was also more appropriate for it, as they should be installed before boot. * Reverted changes to system.cpp Got missed in previous commit, idk why * Silent some compiler warnings * Move sys_process_exit * Add <typeinfo> * Backup logs to old_logs/ * Fix initial log message * Implemented MFF_HIDDEN for VNIDs Minor refactoring Now possible: ```REF_FUNC(...).flag(MFF_HIDDEN);``` * Implemented syscall _sys_process_exit * Restore UTF-8 in Log.cpp * Fixed global variables in Log.cpp * Remove unnecessary newline in Log.cpp * Implemented __sys_look_ctype_table Implemented _sys_tolower, _sys_toupper Fixed _sys_strncasecmp * Implemented _sys_prx_load_module_on_memcontainer * Rewrite _sys_memcmp * Fix _sys_strncasecmp * Rewrite _sys_strlen * Rewrite _sys_strcmp * Rewrite _sys_strncmp * Rewrite _sys_strcat * Rewrite _sys_strcpy * Rewrite _sys_strchr * Rewrite _sys_strrchr * Rewrite _sys_strncpy * Patch: new 'load' syntax [load, path_name1] - load sequence [load, path_name2, -0x40] - same, but modify the addresses by -64 * Rewrite _sys_strncat * Exitspawn support renewal Implemented _sys_process_exit2 syscall Rewritten sys_game_process_exitspawn Rewritten sys_game_process_exitspawn2 Implemented _sys_process_atexitspawn Implemented _sys_process_at_Exitspawn And some other changes * Fix error on loading a library twice * sys_fs_mkdir: implement ENOENT * Try to fix some fs errors * Fix library loading mode 'both' * Remove unnecessary include git-version.h * cellAudio: use error_code * Minor VFS fix * rsx/vk/gl: Refactoring and reimplementation of blit engine Fix rsx offscreen-render-to-display-buffer-blit surface reads - Also, properly scale display output height if reading from compressed tile gl: Fix broken dst height computation - The extra padding is only there to force power-of-2 sizes and isnt used gl: Ignore compression scaling if output is rendered to in a renderpass rsx/gl/vk: Cleanup for GPU texture scaling. Initial impl [WIP] - TODO: Refactor more shared code into RSX/common * rsx/vk: Bug fixes - Make each frame context own its own memory - Fix GPU blit - Fix image layout transitions in flip vk: Improve frame-local memory usage tracking to prevent overwrites - Also slightly bumps VRAM requirements for stream buffers to help with running out of storage - Fixes flickering and missing graphics in some cases. Flickering is still there and needs more work vk: Up vertex attribute heap size and increase the guard size on it vulkan: Reorganize memory management vulkan: blit cleanup vulkan: blit engine improvements - Override existing image mapping when conflicts detected - Allow blitting of depth/stencil surfaces * rsx: Several fixes and improvements - Do not ignore non-centered pixel blitting - Register method ac00+16 - Bump texture memory heap to account for GPU texture scaling requirements (vulkan) - Explicit MRT location index output to better convey intent (openGL) * esx: Fixes to the texture cache rsx: Blit engine improvements - Always handle blits to and from framebuffers through the GPU - Handle depth surfaces properly when using GL - Check for format mismatches when blitting to the surface store [WIP] * rsx/fp: Shader decompiler fixes - Requires proper 2-pass impl rsx/fp: Catch hanging code blocks rsx/fp: Don't pause on scaling error * rsx: Bug fixes and improvements rsx: Try to skip unknown commands without discarding entire cb * rsx: Add support for batched multidraw gl: Fix multidraw [WIP] rsx: Ignore vertex base when data source is generated using arithmetic vk: Check pending flag before doing fence poke vk/gl: Fix for inlined array and immediate draws rsx: Collapse joined draws when batching * rsx: Texture cache fixes and improvments gl/vk/rsx: Refactoring; unify texture cache code gl: Fixups - Removes rsx::gl::texture class and leave gl::texture intact - Simplify texture create and upload mechanisms - Re-enable texture uploads with the new texture cache mechanism rsx: texture cache - check if bit region fits into dst texture before attempting to copy gl/vk: Cleanup - Set initial texture layout to DST_OPTIMAL since it has no data in it anyway at the start - Move structs outside of classes to avoid clutter * gl: cleanup; fix program linkage on mesa using GL_ARB_explicit_uniform_location, also make use of ARB_multidraw * gl: Minor optimizations rsx: Texture cache - improvements to locking rsx: Minor optimizations to get_current_vertex_program and begin-end batch flushes rsx: Optimize texture cache storage - Manages storage in blocks of 16MB rsx/vk/gl: Fix swizzled texture input gl: Hotfix for compressed texture formats * rsx: Fix multidraw range splits again rsx: Hotfix for disjoint range detection * rsx: Texture cache improvements - Limits buffer size to min 720 in the Y axis (1024 section causes conflicts in some cases - TODO) rsx: Fixups to allow large textures for blit operation - Also includes checks for both leaking sections and blit regions for vulkan hotfix for hanging when using WCB addendum - unlock both ro and no blocks before attempting to copy memory blocks gl: Fixups for ARB_explicit_uniform_location - Forces glsl v 430 to make use of the extension rsx/vk: Rework texture cache to minimize recursive access violations - Also modifies the vulkan commandbuffer begin/end/submit mechanism gl: Fix cached_texture_section::is_flushable to take memory protection into account rsx: Fix blit dst offset calculation * rsx: Fixups * rsx: Add support for non-projective shadow sampling - Fixes missing shadows in persona 5 vk: Enable polygon depth bias a.k.a polygonOffset - Fixes shadow acne in persona 5 * rsx: Texture cache fixes and improvements rsx: Conditional lock hack removed vulkan - Fixes - Remove unused texture class - Fix native pitch calculation (WCB) rsx: Catch hanging begin/end pairs when flushing deferred draw calls vulkan: Register DXT compressed formats vulkan: Register depth formats gl: Workaround for 'texture stitching' when gathering flip surface - TODO: Add a proper flip hack option rsx: Fix texture memory size calculation - DXT textures dont have real pitch. Since pitch is used to calculate memory size, make sure it always evaluates to rsx_size rsx: Fix cpu copy detection rsx: Validate blit dst surface and dont make assumptions about region blit order - Also relax restrictions on memory owned by the blit engine if strict rendering is not enabled rsx: Fix depth texture detection rsx: Do not manually offset into dst. The overlapped range check does so automatically rsx: Minor optimizations rsx: Minor fixes - Fix to detect incompatible formats when using GPU texture scaling and show message - Better 'is_depth_texture' algorithm to eliminate false positives * Deploy QT DLLs using windeployqt tool (RPCS3#3432) * Deploy QT DLLs using windeployqt tool Replace the old mechanism that manually copy the DLLs using either CMake or VS built-in functions. The new approach uses the windeployqt tool provided by the QT project that automatically detect the needed DLLs and perform the necessary copying. This approach should be more robust if there's an upstream change on QT project regarding DLLs usage. * QT DLLs: do not ship RPCS3 with ANGLE and software-OpenGL According to https://doc.qt.io/qt-5/windows-requirements.html ANGLE is useful for applications that use OpenGL ES by removing the need to install OpenGL drivers, redirecting OpenGL calls to DirectX instead. RPCS does not uses OpenGL ES so it's better to just remove it from the binary distribution. opengl32sw.dll is a fallback mechanism when ANGLE libraries are not available that uses a software backend for OpenGL. It's unlikely that RPCS3 will work using software-OpenGL, so there's little point into shipping them. * QT DLLs: do not ship with SVG and translation files Like the other removed QT libraries: they are not needed right now. We can deploy them if they become useful someday. * Allow null argument Should fix Jak 2 * gl: Check that fence is not empty before calling destroy * rsx: Fixes - Fix section scanning range for early reject - Specify IMAGE_ASPECT_STENCIL when uploading image_from_cpu * vs/qt: Explicitly reference target output to avoid failing post-build step - Referencing the whole build folder will err if there is any non-qt exe in the folder * Fix branch name for AppVeyor Builds (RPCS3#3472) * rsx: Texture cache fixes - Update section flags when requested - Fix nullptr dereference: cached_dest will be null if dst_is_render_target is true * rsx: Make the 3rd texture dimension matter - Affects cube maps and texture3D surfaces * vulkan: Optimize memory allocation * vulkan: Use frameid to determine texture storage to discard * vk: New images always should be UNDEFINED or PREINITIALIZED * rsx: Clean up debug overlays. Add unreleased textures metric to track texture memory * vk: Clamp unreleased texture metric - TODO: Find the leaking references * ds4: Fix initial connection issue and hotplug/dongle calibration detection * Removed tropdir checks as trophies are already installed I believe trophies to already be installed once the game is launched, according to my dev_hdd0\home\00000001\trophy folder anyway. * Added automatic .pkg installation to games on initial boot * fs: minor fix * sys_fs: use g_tls_error for diagnostic Affected syscalls: sys_fs_open sys_fs_opendir sys_fs_stat sys_fs_mkdir sys_fs_fcntl (get free space) * Qt: Remove duplicate gui repaints in settings dialog ApplyGuiOptions(true) already forces GuiStylesheetRequest and therefore RepaintGui in main_window if necessary. We don't need GuiRepaintRequest on accept because there should be no option left that doesn't repaint immediately * Qt: improve gamelist refresh, fix some more scroll issues despite earlier efforts with modelindex and so on the annoying scroll jump still surfaced sometimes, so we only scrollTo index on selected occasions from now on. Furthermore we don't need to refresh the gamelist if no custom config has been applied, so this will only be needed on accept from now on * Fix buffer overrun in cert loading, made it more accurate. Fixes RPCS3#3408 * sys_mmapper: minor range fix * Improve lv2_memory object * Implement fs::error::notempty * Vulkan: use pkg-config instead of checking for Linux (RPCS3#3440) Makes Vulkan optional on Linux / makes it possible on FreeBSD. * Update Visual Studio 2015 URL * Qt: add char count label to osk dialog * Add git_branch to git-version.h, add display version info to gs_frame and add branch and version to log. (RPCS3#3186) * Added version number and branch name to gs_frame and log file. This also involved making the files that generate git-version.h , get the branch. * Qt: settings_dialog layout fix (hack) * HGA calls sys_rsx_context_attribute before calling sys_rsx_memory_allocate. This inevitably causes RPCS3 to crash. Add nullptr check to prevent this. * Silent some compiler warnings * [Qt] fix graphics adapter selection and rebase for less redundancy * Move sys_process_exit * Add <typeinfo> * Backup logs to old_logs/ * Fix initial log message * Implemented MFF_HIDDEN for VNIDs Minor refactoring Now possible: ```REF_FUNC(...).flag(MFF_HIDDEN);``` * Implemented syscall _sys_process_exit * Restore UTF-8 in Log.cpp * Fixed global variables in Log.cpp * Remove unnecessary newline in Log.cpp * Implemented __sys_look_ctype_table Implemented _sys_tolower, _sys_toupper Fixed _sys_strncasecmp * Implemented _sys_prx_load_module_on_memcontainer * Rewrite _sys_memcmp * Fix _sys_strncasecmp * Rewrite _sys_strlen * Rewrite _sys_strcmp * Rewrite _sys_strncmp * Rewrite _sys_strcat * Rewrite _sys_strcpy * Rewrite _sys_strchr * Rewrite _sys_strrchr * Rewrite _sys_strncpy * Patch: new 'load' syntax [load, path_name1] - load sequence [load, path_name2, -0x40] - same, but modify the addresses by -64 * Rewrite _sys_strncat * Exitspawn support renewal * Fix error on loading a library twice * sys_fs_mkdir: implement ENOENT * rsx/vk/gl: Refactoring and reimplementation of blit engine Fix rsx offscreen-render-to-display-buffer-blit surface reads - Also, properly scale display output height if reading from compressed tile gl: Fix broken dst height computation - The extra padding is only there to force power-of-2 sizes and isnt used gl: Ignore compression scaling if output is rendered to in a renderpass rsx/gl/vk: Cleanup for GPU texture scaling. Initial impl [WIP] - TODO: Refactor more shared code into RSX/common * rsx/vk: Bug fixes - Make each frame context own its own memory - Fix GPU blit - Fix image layout transitions in flip vk: Improve frame-local memory usage tracking to prevent overwrites - Also slightly bumps VRAM requirements for stream buffers to help with running out of storage - Fixes flickering and missing graphics in some cases. Flickering is still there and needs more work vk: Up vertex attribute heap size and increase the guard size on it vulkan: Reorganize memory management vulkan: blit cleanup vulkan: blit engine improvements - Override existing image mapping when conflicts detected - Allow blitting of depth/stencil surfaces * Try to fix some fs errors * Fix library loading mode 'both' * Remove unnecessary include git-version.h * cellAudio: use error_code * Minor VFS fix * rsx: Several fixes and improvements - Do not ignore non-centered pixel blitting - Register method ac00+16 - Bump texture memory heap to account for GPU texture scaling requirements (vulkan) - Explicit MRT location index output to better convey intent (openGL) * esx: Fixes to the texture cache rsx: Blit engine improvements - Always handle blits to and from framebuffers through the GPU - Handle depth surfaces properly when using GL - Check for format mismatches when blitting to the surface store [WIP] * rsx/fp: Shader decompiler fixes - Requires proper 2-pass impl rsx/fp: Catch hanging code blocks rsx/fp: Don't pause on scaling error * rsx: Bug fixes and improvements rsx: Try to skip unknown commands without discarding entire cb * rsx: Add support for batched multidraw gl: Fix multidraw [WIP] rsx: Ignore vertex base when data source is generated using arithmetic vk: Check pending flag before doing fence poke vk/gl: Fix for inlined array and immediate draws rsx: Collapse joined draws when batching * gl: cleanup; fix program linkage on mesa using GL_ARB_explicit_uniform_location, also make use of ARB_multidraw * rsx: Texture cache fixes and improvments gl/vk/rsx: Refactoring; unify texture cache code gl: Fixups - Removes rsx::gl::texture class and leave gl::texture intact - Simplify texture create and upload mechanisms - Re-enable texture uploads with the new texture cache mechanism rsx: texture cache - check if bit region fits into dst texture before attempting to copy gl/vk: Cleanup - Set initial texture layout to DST_OPTIMAL since it has no data in it anyway at the start - Move structs outside of classes to avoid clutter * gl: Minor optimizations rsx: Texture cache - improvements to locking rsx: Minor optimizations to get_current_vertex_program and begin-end batch flushes rsx: Optimize texture cache storage - Manages storage in blocks of 16MB rsx/vk/gl: Fix swizzled texture input gl: Hotfix for compressed texture formats * rsx: Fix multidraw range splits again rsx: Hotfix for disjoint range detection * rsx: Texture cache improvements - Limits buffer size to min 720 in the Y axis (1024 section causes conflicts in some cases - TODO) rsx: Fixups to allow large textures for blit operation - Also includes checks for both leaking sections and blit regions for vulkan hotfix for hanging when using WCB addendum - unlock both ro and no blocks before attempting to copy memory blocks gl: Fixups for ARB_explicit_uniform_location - Forces glsl v 430 to make use of the extension rsx/vk: Rework texture cache to minimize recursive access violations - Also modifies the vulkan commandbuffer begin/end/submit mechanism gl: Fix cached_texture_section::is_flushable to take memory protection into account rsx: Fix blit dst offset calculation * rsx: Fixups * rsx: Add support for non-projective shadow sampling - Fixes missing shadows in persona 5 vk: Enable polygon depth bias a.k.a polygonOffset - Fixes shadow acne in persona 5 * rsx: Texture cache fixes and improvements rsx: Conditional lock hack removed vulkan - Fixes - Remove unused texture class - Fix native pitch calculation (WCB) rsx: Catch hanging begin/end pairs when flushing deferred draw calls vulkan: Register DXT compressed formats vulkan: Register depth formats gl: Workaround for 'texture stitching' when gathering flip surface - TODO: Add a proper flip hack option rsx: Fix texture memory size calculation - DXT textures dont have real pitch. Since pitch is used to calculate memory size, make sure it always evaluates to rsx_size rsx: Fix cpu copy detection rsx: Validate blit dst surface and dont make assumptions about region blit order - Also relax restrictions on memory owned by the blit engine if strict rendering is not enabled rsx: Fix depth texture detection rsx: Do not manually offset into dst. The overlapped range check does so automatically rsx: Minor optimizations rsx: Minor fixes - Fix to detect incompatible formats when using GPU texture scaling and show message - Better 'is_depth_texture' algorithm to eliminate false positives * Allow null argument Should fix Jak 2 * gl: Check that fence is not empty before calling destroy * Deploy QT DLLs using windeployqt tool (RPCS3#3432) * Deploy QT DLLs using windeployqt tool Replace the old mechanism that manually copy the DLLs using either CMake or VS built-in functions. The new approach uses the windeployqt tool provided by the QT project that automatically detect the needed DLLs and perform the necessary copying. This approach should be more robust if there's an upstream change on QT project regarding DLLs usage. * QT DLLs: do not ship RPCS3 with ANGLE and software-OpenGL According to https://doc.qt.io/qt-5/windows-requirements.html ANGLE is useful for applications that use OpenGL ES by removing the need to install OpenGL drivers, redirecting OpenGL calls to DirectX instead. RPCS does not uses OpenGL ES so it's better to just remove it from the binary distribution. opengl32sw.dll is a fallback mechanism when ANGLE libraries are not available that uses a software backend for OpenGL. It's unlikely that RPCS3 will work using software-OpenGL, so there's little point into shipping them. * QT DLLs: do not ship with SVG and translation files Like the other removed QT libraries: they are not needed right now. We can deploy them if they become useful someday. * rsx: Fixes - Fix section scanning range for early reject - Specify IMAGE_ASPECT_STENCIL when uploading image_from_cpu * vs/qt: Explicitly reference target output to avoid failing post-build step - Referencing the whole build folder will err if there is any non-qt exe in the folder * rsx: Texture cache fixes - Update section flags when requested - Fix nullptr dereference: cached_dest will be null if dst_is_render_target is true * Fix branch name for AppVeyor Builds (RPCS3#3472) * ds4: Fix initial connection issue and hotplug/dongle calibration detection * rsx: Make the 3rd texture dimension matter - Affects cube maps and texture3D surfaces * vulkan: Optimize memory allocation * vulkan: Use frameid to determine texture storage to discard * vk: New images always should be UNDEFINED or PREINITIALIZED * rsx: Clean up debug overlays. Add unreleased textures metric to track texture memory * vk: Clamp unreleased texture metric - TODO: Find the leaking references

PeterMcteague changed the title ~~Added version number to gs_frame~~ Added version number to gs_frame and added branch name to gs_frame and log files Aug 16, 2017

Megamouse reviewed Aug 20, 2017

View reviewed changes

PeterMcteague changed the title ~~Added version number to gs_frame and added branch name to gs_frame and log files~~ Add git_branch to git-version.h, add display version info to gs_frame and add branch and version in log Aug 24, 2017

PeterMcteague changed the title ~~Add git_branch to git-version.h, add display version info to gs_frame and add branch and version in log~~ Add git_branch to git-version.h, add display version info to gs_frame and add branch and version to log Aug 24, 2017

hcorion reviewed Aug 26, 2017

View reviewed changes

Added version number and branch name to gs_frame and log file. This a…

11788ba

…lso involved making the files that generate git-version.h , get the branch.

PeterMcteague force-pushed the master branch from a61df3f to 11788ba Compare September 10, 2017 01:10

PeterMcteague changed the title ~~Add git_branch to git-version.h, add display version info to gs_frame and add branch and version to log~~ Add git_branch to git-version.h, add display version info to gs_frame and add branch and version to log. Sep 10, 2017

hcorion approved these changes Sep 13, 2017

View reviewed changes

Nekotekina requested changes Sep 13, 2017

View reviewed changes

PeterMcteague force-pushed the master branch from b703884 to 11788ba Compare September 13, 2017 21:19

PeterMcteague added 2 commits September 13, 2017 22:42

Removed unnecesary ifdefs from git_branch in log and frame

562f379

These aren't neccesary as git_branch will definitely be defined, as it's generated.

Moved branch to rpcs3::version and cleaned up branch code

cdc7e22

-Version for gs_frame is now one line -Unncesary ifdefs removed again

Nekotekina requested changes Sep 15, 2017

View reviewed changes

PeterMcteague and others added 2 commits September 15, 2017 16:12

Changed rpcs3::branch to rpcs3::get_branch()

985dfc3

It's now inside the namespace and is just a function returning a std::string which presents less issues than the previous implementation.

Merge branch 'master' into master

0543f77

Nekotekina approved these changes Sep 16, 2017

View reviewed changes

Nekotekina merged commit c18e71c into RPCS3:master Sep 16, 2017

Nekotekina reviewed Sep 16, 2017

View reviewed changes

AniLeo mentioned this pull request Sep 18, 2017

Add RPCS3 version/build to the game window title. #3152

Closed

Add git_branch to git-version.h, add display version info to gs_frame and add branch and version to log. #3186

Add git_branch to git-version.h, add display version info to gs_frame and add branch and version to log. #3186

Conversation

PeterMcteague commented Aug 9, 2017

Asinin3 commented Aug 10, 2017 • edited Loading

hcorion commented Aug 13, 2017

Asinin3 commented Aug 13, 2017

hcorion commented Aug 13, 2017

PeterMcteague commented Aug 13, 2017 • edited Loading

hcorion commented Aug 13, 2017

PeterMcteague commented Aug 15, 2017 • edited Loading

Asinin3 commented Aug 16, 2017

PeterMcteague commented Aug 16, 2017

Asinin3 commented Aug 16, 2017

PeterMcteague commented Aug 16, 2017 • edited Loading

hcorion commented Aug 18, 2017

Choose a reason for hiding this comment

Choose a reason for hiding this comment

hcorion commented Aug 26, 2017 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

hcorion commented Aug 28, 2017 • edited Loading

PeterMcteague commented Aug 28, 2017 • edited Loading

hcorion commented Aug 28, 2017

PeterMcteague commented Aug 30, 2017

scribam commented Sep 9, 2017

PeterMcteague commented Sep 9, 2017 • edited Loading

hcorion left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

PeterMcteague Sep 13, 2017 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Nekotekina left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Asinin3 commented Aug 10, 2017 •

edited

Loading

PeterMcteague commented Aug 13, 2017 •

edited

Loading

PeterMcteague commented Aug 15, 2017 •

edited

Loading

PeterMcteague commented Aug 16, 2017 •

edited

Loading

hcorion commented Aug 26, 2017 •

edited

Loading

hcorion commented Aug 28, 2017 •

edited

Loading

PeterMcteague commented Aug 28, 2017 •

edited

Loading

PeterMcteague commented Sep 9, 2017 •

edited

Loading

PeterMcteague Sep 13, 2017 •

edited

Loading