From: Jonas Witschel <diabonas@archlinux.org>
Date: Sun, 14 Nov 2021 17:56:49 +0000 (+0100)
Subject: Add ARCHIVE_READDISK_NO_SPARSE to suppress reading sparse file info
X-Git-Tag: v3.6.0~37^2~3
X-Git-Url: http://git.ipfire.org/?a=commitdiff_plain;h=2a8bb4242db773733cca419c8aafd1882c77d75e;p=thirdparty%2Flibarchive.git

Add ARCHIVE_READDISK_NO_SPARSE to suppress reading sparse file info

Sparse file information depends on the file system and can therefore be a
source of unreproducibility for the generated archives, e.g. if the same
content is compressed on a file system with and without sparse file support.
Add an option to suppress reading this information from disk entirely.
---

diff --git a/libarchive/archive.h b/libarchive/archive.h
index fbc96bfe7..9fcb63024 100644
--- a/libarchive/archive.h
+++ b/libarchive/archive.h
@@ -1024,6 +1024,8 @@ __LA_DECL int  archive_read_disk_set_atime_restored(struct archive *);
 #define	ARCHIVE_READDISK_NO_ACL			(0x0020)
 /* Default: File flags are read from disk. */
 #define	ARCHIVE_READDISK_NO_FFLAGS		(0x0040)
+/* Default: Sparse file information is read from disk. */
+#define	ARCHIVE_READDISK_NO_SPARSE		(0x0080)
 
 __LA_DECL int  archive_read_disk_set_behavior(struct archive *,
 		    int flags);
diff --git a/libarchive/archive_read_disk.3 b/libarchive/archive_read_disk.3
index 82d6a5c85..a7e376fff 100644
--- a/libarchive/archive_read_disk.3
+++ b/libarchive/archive_read_disk.3
@@ -137,6 +137,9 @@ for more information on extended file attributes.
 .It Cm ARCHIVE_READDISK_RESTORE_ATIME
 Restore access time of traversed files.
 By default, access time of traversed files is not restored.
+.It Cm ARCHIVE_READDISK_NO_SPARSE
+Do not read sparse file information.
+By default, sparse file information is read from disk.
 .El
 .It Xo
 .Fn archive_read_disk_set_symlink_logical ,
diff --git a/libarchive/archive_read_disk_entry_from_file.c b/libarchive/archive_read_disk_entry_from_file.c
index 9c9cf38ee..ab0270bc2 100644
--- a/libarchive/archive_read_disk_entry_from_file.c
+++ b/libarchive/archive_read_disk_entry_from_file.c
@@ -303,9 +303,11 @@ archive_read_disk_entry_from_file(struct archive *_a,
 		if (r1 < r)
 			r = r1;
 	}
-	r1 = setup_sparse(a, entry, &fd);
-	if (r1 < r)
-		r = r1;
+	if ((a->flags & ARCHIVE_READDISK_NO_SPARSE) == 0) {
+		r1 = setup_sparse(a, entry, &fd);
+		if (r1 < r)
+			r = r1;
+	}
 
 	/* If we opened the file earlier in this function, close it. */
 	if (initial_fd != fd)
diff --git a/libarchive/archive_read_disk_windows.c b/libarchive/archive_read_disk_windows.c
index 877bc449a..a2587c527 100644
--- a/libarchive/archive_read_disk_windows.c
+++ b/libarchive/archive_read_disk_windows.c
@@ -1090,9 +1090,11 @@ next_entry(struct archive_read_disk *a, struct tree *t,
 		}
 
 		/* Find sparse data from the disk. */
-		if (archive_entry_hardlink(entry) == NULL &&
-		    (st->dwFileAttributes & FILE_ATTRIBUTE_SPARSE_FILE) != 0)
-			r = setup_sparse_from_disk(a, entry, t->entry_fh);
+		if ((a->flags & ARCHIVE_READDISK_NO_SPARSE) == 0) {
+			if (archive_entry_hardlink(entry) == NULL &&
+			    (st->dwFileAttributes & FILE_ATTRIBUTE_SPARSE_FILE) != 0)
+				r = setup_sparse_from_disk(a, entry, t->entry_fh);
+		}
 	}
 	return (r);
 }
@@ -2371,9 +2373,11 @@ archive_read_disk_entry_from_file(struct archive *_a,
 		return (ARCHIVE_OK);
 	}
 
-	r = setup_sparse_from_disk(a, entry, h);
-	if (fd < 0)
-		CloseHandle(h);
+	if ((a->flags & ARCHIVE_READDISK_NO_SPARSE) == 0) {
+		r = setup_sparse_from_disk(a, entry, h);
+		if (fd < 0)
+			CloseHandle(h);
+	}
 
 	return (r);
 }