Messages

Does magic zoom work in the build with the timer test for anyone?
http://www.magiclantern.fm/forum/index.php?topic=12627.msg170298#msg170298

This one is to see if the clock I guess still works, testing a stub change for Alex.

That is all up to date, just timer stub changed.

Maybe something compiled wrong with the July 27 version, that was where I first enabled Magic Zoom.

Im going over this issue with my tester, the build in here is not working with Magic Zoom, but his is obviously, let me see what he has enabled at this point compared to forum version.

Thats interesting, I am asking my tester right now to confirm if Magic Zoom was really working for his test.

So original value was right then? It did not look like it was in same spot comparing it to other firmware like 600D and 550D, only reason I changed it.

I could change and post test build for that.

http://www.magiclantern.fm/forum/index.php?topic=12627.msg170329#msg170329

Ok, sorry, here is the updated version, you wont be able to do HDR yet because I will wait for Alex to say good to go on Property Request so you can get exposure override to manually set ISO, only way HDR will work.

Dont Use, buggy with Magic Zoom
https://bitbucket.org/mk11174/1200d_new/downloads/magiclantern-Nightly.2016Jul27.1200D101.zip

Use this instead.
http://www.magiclantern.fm/forum/index.php?topic=12627.msg170329#msg170329

I will post updated version soon since I got ok from alex for those of u that cant compile,  plus repo not updated yet, I will add those changes to it unless alex gets to it before me.

GREAT NEWS!!!

Magic Zoom also confirmed working by my personal tester!

But scared to post without Alex OK!

Also just confirmed Exposure Override worked to get him the manual control of ISO to make HDR work in movie mode because you cant have AUTO ISO when using HDR.

Hi Walter, it seems he is having issues with his card, he did use another card and got 21mb/s like 1100D using benchmark module, but def would not hurt to see if your process will help him, thanks.

Tester Gulwant recorded his first RAW Video on 1200D. Looks like it is setup like the 1100D, was not able to record Raw Video at 1280x720 @24p, it was asking for 36.7mb/s it only did a couple seconds, he had to lower fps to get continuous recording, I think it was 10fps was pretty stable.

He played it back on camera fine too!

Dont want to publish that version until Alex approves though, just wanted to update anyone hoping for 1280x720@24p Raw Video

I have updated link to current test build on will add here so it dont get lost back there.

With the help of tester Gulwant, I coded in the fps values, so ML fps should be normal now.

I also added in the lv buffer addresses from alex, but did not enable anything because Alex has not approved that yet. But screenshot should not record your live view and not scrambled image behind ML overlays.

Also disabled the beep in internals.

Dont use: https://bitbucket.org/mk11174/1200d_new/downloads/magiclantern-Nightly.2016Jul26.1200D101.zip

GOTO Latest here:
http://www.magiclantern.fm/forum/index.php?topic=12627.msg170329#msg170329

yeah, just ignore them, they were compiled, so they snuck in the zip, raw is not enabled yet
so wont work.

I updated ZIP

1200D 1.0.1


In case this is needed here.

should only show 1 meter though as far as i know for mono.

I dont get why it is displaying 2 audio meters here, #define CONFIG_MONO_MIC is set in internals.h?

I will update the build I have in thread. Thanks Alex for catching this.

Everything is updated and in the repo, feature.h is copied mostly from 1100d for now until chart is made so we know whats good or bad, go look and see whats undefined from all features.

Some features you cant see until propert request can be defined, cant do that until port is tested.

And things like magic zoom probably don't work cause I am probably missing a define for the camera somewhere and its registers needed, just like why hdr won't work yet either.

I would need to be refreshed on what is needed to be updated to turn those into working features, I thought it was something with Edmac address,  have not messed with thats since 550d.

https://bitbucket.org/hudson/magic-lantern/commits/a1fa773fa8d7e88ad18a23676ce055cb1de5d448?at=1200D

Also does your camera have stereo mic, or mono mic input like 1100d, its set in internals for mono mic for now since I dont know what hardware your camera has.

UPDATE! I see your camera dont have a mic input and the internal mic is mono, not stereo, so the setting in internals.h is fine then, just like 1100D.

I will always be here to help when I can with the port, so nothing has stopped, I will be busy starting next month when school starts, but I will find time to help. I think I went as far as I could in QEMU, things need to be physically tested on camera and a chart needs to be made about whats buggy, what works, what dont work, thats most important step.

Heres the build of latest, but its new port, so play safe. Any weird things happen, turn off camera and take out battery and restart and report EXACT procedure that caused it to happen in as much detail as possible.

https://bitbucket.org/mk11174/1200d_new/downloads/magiclantern-Nightly.2016Jul24.1200D101.zip

Use this to work with safer build without Magic Zoom for now.
https://bitbucket.org/mk11174/1200d_new/downloads/magiclantern-Nightly.2016Jul26.1200D101.zip

Don't use above, go to latest here: http://www.magiclantern.fm/forum/index.php?topic=12627.msg170329#msg170329

Just make sure to update to Canon 1.0.1 firmware if not already, and install FIR from alex links above to get bootflag going and make card bootable using EOSCard utilty: http://pel.hu/eoscard/  also link posted by alex above of ML install guide.

Updated repository with changes of what we have so far, in case anyone is interested in compiling and building themselves for testing stuff.

Makes sure to report things that work or don't work, because it's still a guessing game. I don't have this camera, so I can't test myself.

But good news is, it works in QEMU and also confirmed working on the actual camera by "gulwantgill"

No RAW VIDEO until the testing phase is complete.

Compile the selftest module and run tests from there, especially Stubs API test.

https://bitbucket.org/hudson/magic-lantern/commits/a1fa773fa8d7e88ad18a23676ce055cb1de5d448?at=1200D

Loaded non RAW features and 2 modules, working fine in QEMU, no more crash so far!

Gotta love Alex and his QEMU with debug log!!!!

NSTUB(0xFF127ACC,  FreeMemoryResource)  WAS  NSTUB(0xFF127AAC,  FreeMemoryResource)

Ok, worth a shot!
And thanks, as usual! lol

Another Issue, most likely related to this crash, is when I go to view memory from config menu, it crashes, it will stay for a few seconds, but does crash and never finds any SRM, shoot contig, or shoot total. Maybe cause of the crash as I enter the Memory menu. In qemu I can use it for a bit, but I can see it crashing in the terminal. This is all replicated on the real camera as well.

Def some sort of memory issue, with or without Allocate_Memory_Pool, no features used except Free_Memory, everything to minimum.

Code Select Expand

[   CtrlSrv:ff44ff18 ] (83:02) DlgShootOlc.c LOCAL_DIALOG_REFRESH
[ guess_mem:00c818f8 ] task_create(stack_try, prio=1e, stack=82800, entry=c813a4, arg=82800)
[   CtrlSrv:ff44ff18 ] (83:02) DlgShootOlc.c LOCAL_DIALOG_REFRESH
[ guess_mem:00c818f8 ] task_create(stack_try, prio=1e, stack=82c00, entry=c813a4, arg=82c00)
[    RscMgr:ff0d8ebc ] (80:02) srmEventDispatch: Current = 0, dwEventID = 0, dwParam = 15 S
[    RscMgr:ff183af0 ] (80:01) AllocateMEM3 15 80
[    RscMgr:ff17eb18 ] (80:01) ###### AllocateMemoryFromShootMemoryObject 8388608
[    RscMgr:ff17ebf4 ] (80:01) ###### 0 0x84305C 0x84305C
[    RscMgr:ff17ebf4 ] (80:01) ###### 1 0x843064 0x842F64
[    RscMgr:ff17ec1c ] (80:01) ###### 1 pMemoryUnit 0x842F64 0x86A60
[    RscMgr:ff17ec78 ] (80:01) AllocatableSizeOfPackHeap 32505728, 16384, 8388608
[    RscMgr:ff17eee0 ] (80:01) !!AllocateMemory 8388608
[    RscMgr:ff17ee8c ] (80:01) AllocateMemoryFromShootMemoryObject 1 78675456 8388608 0x000c48e8
[    RscMgr:ff17eea8 ] (80:01) ###### VirtualXXXFreeSize1 78675456 1866960
[    RscMgr:ff17f0f4 ] (80:01) ###### VirtualXXXFreeSize2 70286832 1866960
[    RscMgr:ff17f108 ] (80:01) ###### !! AllocateMemoryFromShootMemoryObject hMemSuite 0xC48E8
[    RscMgr:ff17f758 ] (80:01) ###### ShrinkMemoryToShootMemoryObject 0xC48E8 0x800000 0x800000
[    RscMgr:ff1841fc ] (80:01), OK AllocDev[2] C48E8 800000 50 1 44CFAC0 44CFAC0 4CCFAD0
[    RscMgr:ff0d90c4 ] (80:02) srmEventDispatch: Current = 0, dwEventID = 0, dwParam = 15 E
[    RscMgr:ff0d8ebc ] (80:02) srmEventDispatch: Current = 0, dwEventID = 0, dwParam = 15 S
[    RscMgr:ff183af0 ] (80:01) AllocateMEM3 15 80
[    RscMgr:ff17eb18 ] (80:01) ###### AllocateMemoryFromShootMemoryObject 4194304
[    RscMgr:ff17ebf4 ] (80:01) ###### 0 0x84305C 0x842F64
[    RscMgr:ff17ec1c ] (80:01) ###### 0 pMemoryUnit 0x842F64 0x86A60
[    RscMgr:ff17ec78 ] (80:01) AllocatableSizeOfPackHeap 24117104, 16384, 4194304
[    RscMgr:ff17eee0 ] (80:01) !!AllocateMemory 4194304
[    RscMgr:ff17ee8c ] (80:01) AllocateMemoryFromShootMemoryObject 1 70286832 4194304 0x000c4948
[    RscMgr:ff17eea8 ] (80:01) ###### VirtualXXXFreeSize1 70286832 1866960
[    RscMgr:ff17f0f4 ] (80:01) ###### VirtualXXXFreeSize2 66092512 1866960
[    RscMgr:ff17f108 ] (80:01) ###### !! AllocateMemoryFromShootMemoryObject hMemSuite 0xC4948
[    RscMgr:ff17f758 ] (80:01) ###### ShrinkMemoryToShootMemoryObject 0xC4948 0x400000 0x400000
[    RscMgr:ff1841fc ] (80:01), OK AllocDev[2] C4948 400000 50 1 40CFAB0 40CFAB0 4CCFAD0
[    RscMgr:ff0d90c4 ] (80:02) srmEventDispatch: Current = 0, dwEventID = 0, dwParam = 15 E
[ guess_mem:ff0c1d60 ] (8b:16) ###exceptionhandlercbr 0x0 0
[ guess_mem:ff0c1e90 ] (8b:16) #####exceptionhandlercbr 0x0
[ guess_mem:ff0c1ea0 ] (8b:03) < Error Exception>

[ guess_mem:ff0c1eb4 ] (8b:03) TYPE        : 4

[ guess_mem:ff0c1ec8 ] (8b:03) ISR         : 0

[ guess_mem:ff0c1edc ] (8b:03) TASK IDSR   : 51445849

[ guess_mem:ff0c1ef0 ] (8b:03) TASK Name   : guess_mem

[ guess_mem:ff0c1f04 ] (8b:03) R 0         : c4948

[ guess_mem:ff0c1f18 ] (8b:03) R 1         : 400000

[ guess_mem:ff0c1f2c ] (8b:03) R 2         : c48e8

[ guess_mem:ff0c1f40 ] (8b:03) R 3         : 0

[ guess_mem:ff0c1f54 ] (8b:03) R 4         : 0

[ guess_mem:ff0c1f68 ] (8b:03) R 5         : cbe570

[ guess_mem:ff0c1f7c ] (8b:03) R 6         : 400000

[ guess_mem:ff0c1f90 ] (8b:03) R 7         : c8c944

[ guess_mem:ff0c1fa4 ] (8b:03) R 8         : 0

[ guess_mem:ff0c1fb8 ] (8b:03) R 9         : 19980218

[ guess_mem:ff0c1fcc ] (8b:03) R10         : 19980218

[ guess_mem:ff0c1fe0 ] (8b:03) R11         : 19980218

[ guess_mem:ff0c1ff4 ] (8b:03) R12         : 0

[ guess_mem:ff0c2008 ] (8b:03) R13         : 1603c

[ guess_mem:ff0c2244 ] (8b:03) R14         : 0

[ guess_mem:ff0c2258 ] (8b:03) PC          : 0

[ guess_mem:ff0c226c ] (8b:03) CPSR        : 20000093

[ guess_mem:ff0c22b0 ] (8b:16), Exception : Time 2000/1/1 0:0:31
[   CtrlSrv:ff44ff18 ] (83:02) DlgShootOlc.c LOCAL_DIALOG_REFRESH
[    DbgMgr:ff125eb0 ] (00:01) [PM] DisablePowerSave (Counter = 3)

Tracked Save Config Now crash down to fio_free(msg) in config.c
here: and if I comment it out in these 2 places depending if I have a cfg file already on card, no crash when I use Save config now, but not sure where to look from there, gets a bit confusing?

Code Select Expand

int config_save_file(const char *filename)
{
    int count = 0;

    DebugMsg( DM_MAGIC, 3, "%s: saving to %s", __func__, filename );
   
    #define MAX_SIZE 10240
    char* msg = fio_malloc(MAX_SIZE);
    msg[0] = '\0';
 
    snprintf( msg, MAX_SIZE,
        "# Magic Lantern %s (%s)\n"
        "# Built on %s by %s\n",
        build_version,
        build_id,
        build_date,
        build_user
    );

    struct tm now;
    LoadCalendarFromRTC( &now );

    snprintf(msg + strlen(msg), MAX_SIZE - strlen(msg),
        "# Configuration saved on %04d/%02d/%02d %02d:%02d:%02d\n",
        now.tm_year + 1900,
        now.tm_mon + 1,
        now.tm_mday,
        now.tm_hour,
        now.tm_min,
        now.tm_sec
    );

    for(struct config_var *var = _config_vars_start; var < _config_vars_end ; var++ )
    {
        if (*(int*)var->value == var->default_value)
            continue;

        snprintf(msg + strlen(msg), MAX_SIZE - strlen(msg) - 1,
            "%s = %d\r\n",
            var->name,
            *(int*) var->value
        );

        count++;
    }
   
    FILE * file = FIO_CreateFile( filename );
    if(!file)
    {
        //fio_free(msg);
        return -1;
    }
   
    FIO_WriteFile(file, msg, strlen(msg));

    FIO_CloseFile( file );
   
   //fio_free(msg)
   
    return count;
}

I double checked all /** File I/O **/ stubs, and /** Memory allocation **/ stubs, all were fine.


UPDATE!!!!, that MAX_SIZE caught my eye, I figured since I am in QEMU, no big deal playing, I changed it to 1024, no more crash?

Well, I was able to get menu back, I guess one of the crashes screwed the mounted sd image cause the files would not unlock.

Anyways, I did as you said with the AUTOSAVE.NEG and it fixed the crash leaving the menu.

So, since you brought up saving the config file, I went right to Config menu and click Save Config now, and it crashed right away.