Mihir Luthra

> Can you add something to that spreadsheet to show the percentage speedup/slowdown? I have added a new column indicating the % speedup/slowdown. Also I guess there are still many...

> Hrm. Looks like all this is a lot of work for only a few percent change on average... Till now I guess yes. I am hoping to see more...

@pmetzger I have made some changes to the % calculation formula. Last time it was getting calculated as `((actual_time_by_old_trace_mode - actual_time_by_modified_trace_mode)/actual_time_by_old_trace_mode) * 100` which actually isn’t correct. I changed it...

> 8.5 minutes vs 7.25 minutes is not a 51% inprovement. A 51% improvement would be 4.25 minutes. I understand, I will try to make the improvements better. Maybe I...

> > 8.5 minutes vs 7.25 minutes is not a 51% inprovement. A 51% improvement would be 4.25 minutes. > > ... only if the build without trace mode would...

@raimue Kindly check the latest code. I think I fixed most issues which you told me last time. Also, [here](https://docs.google.com/spreadsheets/d/1ksj3Fex-AnTEU4f4IRzwUkTpN4XfUye-HqSdZwXOsKs/edit#gid=0) are the latest tests but they still are incomplete.

I thought I should update it here once, the shared memory to be used alongside the above [ctrie repository](https://github.com/neverpanic/cctrie) is ready: https://github.com/MihirLuthra/shm_alloc Thanks to Clemens, the combination is significantly faster...

Old changes have been archived : [old branch](https://github.com/MihirLuthra/macports-base/tree/dtsm-darwintrace-archive) **Additions**: 1. [Shared memory allocation library](https://github.com/MihirLuthra/shm_alloc). This is introduced into the memory of each darwintraced process with`__attribute__((constructor))`. 2. [Ctrie data structure working...

Yes, this solves the issue (I can confirm for `armv7-linux-androideabi`) @thomcc, pinging you since you seem to be an active reviewer of `cmake-rs` 🙂. Please ignore if this PR falls...