On 4/29/2014 8:43 AM, Sebastian Huber wrote: > On 2014-04-29 15:32, Joel Sherrill wrote: >> Are you sure the code which reads the cause register and vectors is not >> used? I >> am off today but my recollection is that without that code, no isr is going >> to >> get vectored. >> >> The MIPS already required BSP/CPU model specific deciding of isr source. The >> cause decoding has to come from somewhere and I am positive we preserved it >> in >> the transition from simple to PIC. >> > These vectorisrs.c files in libcpu didn't show up in any Makefile.am. It > seems > someone copied them to libbsp and worked with these copies: > > ./c/src/lib/libbsp/mips/jmr3904/irq/vectorisrs.c > ./c/src/lib/libbsp/mips/rbtx4925/irq/vectorisrs.c > ./c/src/lib/libbsp/mips/hurricane/irq/vectorisrs.c > ./c/src/lib/libbsp/mips/malta/irq/vectorisrs.c > ./c/src/lib/libbsp/mips/csb350/irq/vectorisrs.c > ./c/src/lib/libbsp/mips/genmongoosev/irq/vectorisrs.c > ./c/src/lib/libbsp/mips/rbtx4938/irq/vectorisrs.c > ./c/src/lib/libcpu/mips/tx49/vectorisrs/vectorisrs.c > ./c/src/lib/libcpu/mips/r46xx/vectorisrs/vectorisrs.c > ./c/src/lib/libcpu/mips/tx39/vectorisrs/vectorisrs.c > ./c/src/lib/libcpu/mips/au1x00/vectorisrs/vectorisrs.c > ./c/src/lib/libcpu/mips/rm52xx/vectorisrs/vectorisrs.c > ./c/src/lib/libcpu/mips/mongoosev/vectorisrs/vectorisrs.c > OK. These files were migrated to the BSPs that actually use them. And updated as needed.
They are dead. I have wondered about dispersing the contents of libcpu to libbsp/shared and libbsp/. I suspect significant portions of it are only used by a single BSP and could be part of that BSP. Others would just move to libbsp/shared or libbsp/CPU/shared. At least the portions that are in a single BSP would make a nice dent in it. -- Joel Sherrill, Ph.D. Director of Research & Development joel.sherr...@oarcorp.com On-Line Applications Research Ask me about RTEMS: a free RTOS Huntsville AL 35805 Support Available (256) 722-9985 _______________________________________________ rtems-devel mailing list rtems-devel@rtems.org http://www.rtems.org/mailman/listinfo/rtems-devel