| 1 |
On 13/07/2020 00:57, Ahmed Soliman wrote: |
| 2 |
> I tried it on the real gdb code base and unfortunately it crashed but |
| 3 |
> due to completely different reasons (early crash). I did in fact skip |
| 4 |
> multiple details, I was inserting printlns everywhere and hacking |
| 5 |
> through the code trying to just get it working. There were things (for |
| 6 |
> example : mmap ) disabled in dlmalloc but was required for |
| 7 |
> for mspaces to work. |
| 8 |
|
| 9 |
Keep in mind that println/fmt may allocate. |
| 10 |
|
| 11 |
> Now that I know that the basic idea worked for the memory corruption |
| 12 |
> proof of concept, I would just git reset --hard and try to implement |
| 13 |
> the fix as clean as possible. |
| 14 |
> |
| 15 |
> What slightly worries me is that I could be totally unlucky and the |
| 16 |
> crash in gcc and segfault in the POC I created end up being completely |
| 17 |
> different bugs, and I will have to restart wondering what is causing |
| 18 |
> memory corruptions. |
| 19 |
> |
| 20 |
|
| 21 |
You are fixing bugs that are valid and must be fixed as soon as |
| 22 |
possible, so do not worry, progress is progress :) |
| 23 |
|
| 24 |
On the other hand, I wonder if upstream gcc doesn't have a mean to avoid |
| 25 |
completely dlopen and dynamic linking. |
| 26 |
|
| 27 |
lu |
Archives