Re: [PATCH v2 3/9] scsi: build scsi_common.o for all scsi passthrough request users

From: Bart Van Assche
Date: Tue Jul 31 2018 - 16:01:29 EST


On Tue, 2018-07-31 at 12:51 -0700, Kees Cook wrote:
+AD4- diff --git a/drivers/Makefile b/drivers/Makefile
+AD4- index 24cd47014657..a6abd7a856c6 100644
+AD4- --- a/drivers/Makefile
+AD4- +-+-+- b/drivers/Makefile
+AD4- +AEAAQA- -76,7 +-76,7 +AEAAQA- obj-+ACQ-(CONFIG+AF8-DMA+AF8-SHARED+AF8-BUFFER) +-+AD0- dma-buf/
+AD4- obj-+ACQ-(CONFIG+AF8-NUBUS) +-+AD0- nubus/
+AD4- obj-y +-+AD0- macintosh/
+AD4- obj-+ACQ-(CONFIG+AF8-IDE) +-+AD0- ide/
+AD4- -obj-+ACQ-(CONFIG+AF8-SCSI) +-+AD0- scsi/
+AD4- +-obj-y +-+AD0- scsi/
+AD4- obj-y +-+AD0- nvme/
+AD4- obj-+ACQ-(CONFIG+AF8-ATA) +-+AD0- ata/
+AD4- obj-+ACQ-(CONFIG+AF8-TARGET+AF8-CORE) +-+AD0- target/

The above change not only selects scsi+AF8-common.o but also the following object
files: scsi.o hosts.o scsi+AF8-ioctl.o scsicam.o scsi+AF8-error.o scsi+AF8-lib.o scsi+AF8-scan.o
scsi+AF8-sysfs.o scsi+AF8-devinfo.o scsi+AF8-trace.o scsi+AF8-logging.o. That is a change that
has not been mentioned in the patch description. That makes me wonder whether
this is an intended change? Wouldn't a cleaner solution have been to move
scsi+AF8-common.c into a new directory?

Bart.