Hi Andrew, I can speak to the missing aci.h file. As you suggested, this header file is property of ADIC (now Quantum) and as such cannot be open sourced. There's a number of files like this in the device code and in the file system code. The second drop of SAM-QFS open source code will have the necessary changes to build around these missing header files. Until then, the SAM-QFS open source code is not buildable.
Even once buildable, SAM will only work with direct attached libraries because of the missing header files for the ADIC, Sony, IBM3494 and ACSLS library code. Mary Andrew C. Henle wrote: > I tried to build the SAM-QFS source released a few months ago. There appear > to be fatal flaws in the package as released. > > I'm building on a stock install of snv_89. > > First, and perhaps the most important, the "san.c" file is missing from the > "samfs" kernel module directory. The Build.mk makefile specifies it as part > of the build, but it's not there. Now, it may be possible for the module to > work properly without the missing code, but I'm skeptical that a SAN file > system module missing source code from a file called "san.c" could safely be > considered viable, although the "samfs" module does build. > > Second, the "aci.h" file is missing. From the failed references, I suspect > it's from ADIC. But without it, the build fails to build a lot of utilities > necessary to successfully run SAM-QFS. And the missing information is far too > extensive to do without. > > It's possible other issues are lurking, but without the "aci.h" file, I > really can't get any further. > > There are also some smaller problems that can be worked around. The biggest: > two missing header files, "samsanergy/fsmdc.h" and "samsanergy/fsmdcsam.h" > cause some issues. I suspect those files contain IBM/Tivoli-proprietary data, > and their absence isn't difficult to work around as only a few typedef's are > missing, and since the complex types are never dereferenced in the SAM-QFS > code (at least as far as I've gotten into the build, anyway), just treating > the pointers as opaque void pointers does work. I assume as long as I'm not > trying to interface with SANergy, any home-compiled version of SAM-QFS would > be fine without those headers. > > > This message posted from opensolaris.org > _______________________________________________ > storage-discuss mailing list > [email protected] > http://mail.opensolaris.org/mailman/listinfo/storage-discuss -- Mary Matejcek | 1270 Eagan Industrial Rd Sun Microsystems | Eagan, MN 55121 651-554-1527 | Fax: 651-554-1540 _______________________________________________ storage-discuss mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/storage-discuss
