GitLab

see #223

this allows us to request memory allocations at smaller granularity
than MiB

- untested

Anonymous mmap() calls have no backing file and therefore won't match
any Darshan characterization records, so we can skip the record lookup
in this case.  This also fixes a deadlock on Cray systems with hugepages
enabled, see #210.

Signed-off-by: Phil Carns <carns@mcs.anl.gov>

no instrumenation_disabled flags are needed since this
functionality will be managed by the overlying MPI-IO and POSIX
modules (DXT calls are embedded in wrappers in these modules).

Signed-off-by: Cong Xu <cong.xu@intel.com>

Signed-off-by: Cong Xu <cong.xu@intel.com>

Signed-off-by: Cong Xu <cong.xu@intel.com>

2. Refactor DXLT to be a generic IO tracing tool for files on any file system.
3. Add file type information in DXLT log parser.

Next steps:
1. Based on the file type information in DXLT log parser, add information for specific file types. (For instance, add OST information for files on Lustre)
2. In the process of parsing DXLT logs, figure out how to obtain Lustre striping information from Lustre module.
3. Rename DXLT to be "Darshan eXtended Tracing (DXT)" or "Darshan eXtended IO Tracing (DXIOT)".
Signed-off-by: Cong Xu <cong.xu@intel.com>

Signed-off-by: Cong Xu <cong.xu@intel.com>

1. Use Lustre Module in Darshan
2. Integrate DXLT IO tracing code into one module
3. Add hooks in POSIX and MPI Modules
4. Add Intel copyright
Signed-off-by: Cong Xu <cong.xu@intel.com>

The MAP_OR_FAIL macro is resolving fopen rather than fopen64,
causing a NULL function pointer to be executed in the wrapper
in the case of using Darshan's LD_PRELOAD mechanism.