The diff looks good for me. I’ll recompile system with your diff
tomorrow.
> On 16 Aug 2020, at 14:35, Visa Hankala <v...@openbsd.org> wrote:
>
> The msg_bufl field of struct msgbuf is written but never read. The value
> was used by kernfs which is no longer present, so the code could be
> cleaned up a little by removing the field.
>
> On some systems the message buffer data are preserved across a reboot.
> However, the preservation is best-effort only, and initmsgbuf() refuses
> to use the old data if the struct's size has changed.
>
> Changing the struct affects at least dmesg(8). The program has to be
> recompiled.
>
> OK?
>
> Index: kern/subr_log.c
> ===================================================================
> RCS file: src/sys/kern/subr_log.c,v
> retrieving revision 1.66
> diff -u -p -r1.66 subr_log.c
> --- kern/subr_log.c 7 Apr 2020 13:27:51 -0000 1.66
> +++ kern/subr_log.c 16 Aug 2020 10:57:59 -0000
> @@ -151,7 +151,6 @@ msgbuf_putchar(struct msgbuf *mbp, const
>
> s = splhigh();
> mbp->msg_bufc[mbp->msg_bufx++] = c;
> - mbp->msg_bufl = lmin(mbp->msg_bufl+1, mbp->msg_bufs);
> if (mbp->msg_bufx < 0 || mbp->msg_bufx >= mbp->msg_bufs)
> mbp->msg_bufx = 0;
> /* If the buffer is full, keep the most recent data. */
> Index: sys/msgbuf.h
> ===================================================================
> RCS file: src/sys/sys/msgbuf.h,v
> retrieving revision 1.11
> diff -u -p -r1.11 msgbuf.h
> --- sys/msgbuf.h 23 Jun 2016 13:15:21 -0000 1.11
> +++ sys/msgbuf.h 16 Aug 2020 10:57:59 -0000
> @@ -38,7 +38,6 @@ struct msgbuf {
> long msg_bufx; /* write pointer */
> long msg_bufr; /* read pointer */
> long msg_bufs; /* real msg_bufc size (bytes) */
> - long msg_bufl; /* # chars, <= msg_bufs */
> long msg_bufd; /* number of dropped bytes */
> char msg_bufc[1]; /* buffer */
> };
>
