Re: git: fb53e7adaf5b - main - tarfs: Fix 32-bit build.
- In reply to: Dimitry Andric : "Re: git: fb53e7adaf5b - main - tarfs: Fix 32-bit build."
- Go to: [ bottom of page ] [ top of archives ] [ this month ]
Date: Thu, 02 Feb 2023 23:08:50 UTC
On 2 Feb 2023, at 23:04, Dimitry Andric <dim@FreeBSD.org> wrote: > > On 2 Feb 2023, at 21:39, Dag-Erling Smørgrav <des@FreeBSD.org> wrote: >> >> The branch main has been updated by des: >> >> URL: https://cgit.FreeBSD.org/src/commit/?id=fb53e7adaf5b45ca08f650babc40077deff5e9af >> >> commit fb53e7adaf5b45ca08f650babc40077deff5e9af >> Author: Dag-Erling Smørgrav <des@FreeBSD.org> >> AuthorDate: 2023-02-02 20:36:01 +0000 >> Commit: Dag-Erling Smørgrav <des@FreeBSD.org> >> CommitDate: 2023-02-02 20:39:10 +0000 >> >> tarfs: Fix 32-bit build. >> >> Sponsored by: Juniper Networks, Inc. >> Sponsored by: Klara, Inc. > ... >> --- a/sys/fs/tarfs/tarfs_subr.c >> +++ b/sys/fs/tarfs/tarfs_subr.c >> @@ -137,7 +137,7 @@ tarfs_print_node(struct tarfs_node *tnp) >> printf("\tvnode %p\n", tnp->vnode); >> printf("\ttmp %p\n", tnp->tmp); >> printf("\ttype %d\n", tnp->type); >> - printf("\tino %lu\n", tnp->ino); >> + printf("\tino %lu\n", (unsigned long)tnp->ino); >> printf("\tsize %zu\n", tnp->size); >> printf("\tname %s\n", >> (tnp->name == NULL) ? "<<root>>" : tnp->name); >> @@ -146,7 +146,7 @@ tarfs_print_node(struct tarfs_node *tnp) >> printf("\tgid %d\n", tnp->gid); >> printf("\tmode o%o\n", tnp->mode); >> printf("\tflags %u\n", tnp->flags); >> - printf("\tnlink %lu\n", tnp->nlink); >> + printf("\tnlink %lu\n", (unsigned long)tnp->nlink); >> printf("\tatime %d\n", (int)tnp->atime.tv_sec); >> printf("\tmtime %d\n", (int)tnp->mtime.tv_sec); >> printf("\tctime %d\n", (int)tnp->ctime.tv_sec); > > This be just nitpicking, but the ino and nlink fields are typically > defined as int64_t and uint64_t, so these should be printed as such. > > Since PRId64 and PRIu64 are apparently frowned upon, the usual solution > is to cast everything to intmax_t or uintmax_t, and use %j for the > modifier. The casts of what I assume are time_t to int are also concerning. Jess