On Fri, 2007-06-01 at 09:07 -0400, James Bottomley wrote: > On Thu, 2007-05-31 at 16:56 -0700, Randy Dunlap wrote: > > On Thu, 31 May 2007 18:22:55 -0400 James Bottomley wrote: > > > > > On Thu, 2007-05-31 at 18:35 +0200, Toralf Förster wrote: > > > > Hello, > > > > > > > > the build with the attached .config failed, make ends with: > > > > ... > > > > CC arch/i386/boot/compresse > > > > HOSTCC arch/i386/boot/tools/build > > > > BUILD arch/i386/boot/bzImage > > > > Root device is (3, 8) > > > > Boot sector 512 bytes. > > > > Setup is 7178 bytes. > > > > System is 2513 kB > > > > Kernel: arch/i386/boot/bzImage is ready (#1) > > > > Building modules, stage 2. > > > > MODPOST 363 modules > > > > ERROR: "spi_dv_device" [drivers/scsi/aic7xxx/aic7xxx.ko] undefined! > > > > ERROR: "spi_attach_transport" [drivers/scsi/aic7xxx/aic7xxx.ko] undefined! > > > > ERROR: "spi_populate_ppr_msg" [drivers/scsi/aic7xxx/aic7xxx.ko] undefined! > > > > ERROR: "spi_populate_width_msg" [drivers/scsi/aic7xxx/aic7xxx.ko] undefined! > > > > ERROR: "spi_release_transport" [drivers/scsi/aic7xxx/aic7xxx.ko] undefined! > > > > ERROR: "spi_display_xfer_agreement" [drivers/scsi/aic7xxx/aic7xxx.ko] undefined! > > > > ERROR: "spi_populate_sync_msg" [drivers/scsi/aic7xxx/aic7xxx.ko] undefined! > > > > ERROR: "spi_dv_device" [drivers/scsi/53c700.ko] undefined! > > > > ERROR: "spi_attach_transport" [drivers/scsi/53c700.ko] undefined! > > > > ERROR: "spi_schedule_dv_device" [drivers/scsi/53c700.ko] undefined! > > > > ERROR: "spi_release_transport" [drivers/scsi/53c700.ko] undefined! > > > > ERROR: "spi_display_xfer_agreement" [drivers/scsi/53c700.ko] undefined! > > > > ERROR: "spi_populate_sync_msg" [drivers/scsi/53c700.ko] undefined! > > > > ERROR: "spi_print_msg" [drivers/scsi/53c700.ko] undefined! > > > > ERROR: "ROOT_DEV" [drivers/mtd/maps/nettel.ko] undefined! > > > > make[1]: *** [__modpost] Error 1 > > > > make: *** [modules] Error 2 > > > > > > This looks like some build issue. The symbols (except the last > > > ROOT_DEV) are defined in scsi_transport_spi.o which should have been > > > compiled in your build log because your .config has > > > CONFIG_SCSI_SPI_ATTRS=y which should build it. > > > > Easily reproducible. and scsi_transport_spi.o is compiled. > > but apparently not linked. I suppose that it's due to > > CONFIG_SCSI=m, but we still have: > > > > # > > # SCSI Transports > > # > > CONFIG_SCSI_SPI_ATTRS=y > > CONFIG_SCSI_FC_ATTRS=m > > CONFIG_SCSI_ISCSI_ATTRS=m > > CONFIG_SCSI_SAS_ATTRS=m > > CONFIG_SCSI_SAS_LIBSAS=m > > > > Should these modules be constrained to =m when CONFIG_SCSI=m ? > > Yes, they should. If there's no SCSI module there's nothing to link > them into. I thought the depends SCSI directive enforced this > requirement (as in you can't make SCSI_SPI_ATTRS y if SCSI is m and we > depend on it) ... is this incorrect? I just tried it on my setup. When SCSI=m this is what I get: * * SCSI Transports * Parallel SCSI (SPI) Transport Attributes (SCSI_SPI_ATTRS) [M/y/?] (NEW) Also, all the SCSI dependent device drivers are doing this too ... I'm sure this didn't used to be the case, so I suspect we have a bug in Kconfig somewhere. James - To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
