linux-kernel - Re: vmsplice can't work well

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <20060829140542.GN12257@kernel.dk>
Date:	Tue, 29 Aug 2006 16:05:43 +0200
From:	Jens Axboe <axboe@...nel.dk>
To:	Yi Yang <yang.y.yi@...il.com>
Cc:	torvalds@...l.org, linux-kernel@...r.kernel.org
Subject: Re: vmsplice can't work well

On Tue, Aug 29 2006, Yi Yang wrote:
> Hi, Jens
> 
> I try to trace vmsplice and find it can't work in both ppc64 and i386,
> it always return -EFAULT because of the address of iovec.iov_base no
> matter it is page alignment or not, I don't know if I should file a
> bug for it, do you test it on i386 or ppc64?

Please provide an strace of the problem, it works fine for me (on x86-64
and x86, I've previously also tested ppc64 and ia64). Also please see
the splice tools here for more examples:

http://brick.kernel.dk/snaps/splice-git-20060711102502.tar.gz

I patched your program to fix the x86-64 syscall number and one/two
bugs, diff attached. Output for me:

axboe@...son:/home/axboe $ ./f | cat > /dev/null
getpagesize = 4096
page size: 4096 bytes
written len = 4096

--- f.c~	2006-08-29 16:02:21.000000000 +0200
+++ f.c	2006-08-29 16:04:41.000000000 +0200
@@ -22,7 +22,7 @@
 #elif defined(__x86_64__)
 #define __NR_splice 275
 #define __NR_tee 276
-#define __NR_vmsplice 277
+#define __NR_vmsplice 278
 #elif defined(__powerpc__) || defined(__powerpc64__)
 #define __NR_splice 283
 #define __NR_tee 284
@@ -71,23 +71,26 @@
 	v.iov_base = buffer;
 	v.iov_len = len;
 
-	while (len) {
+	while (v.iov_len) {
 		/*
 		 * in a real app you'd be more clever with poll of course,
 		 * here we are basically just blocking on output room and
 		 * not using the free time for anything interesting.
 		 */
 		if (poll(&pfd, 1, -1) < 0)
-		return xerror("poll");
+			return xerror("poll");
 
 		written = vmsplice(fd, &v, 1, 0);
 		printf("here: len = %d, written = %d\n", len, written);
 
-		if (written <= 0)
+		if (!written)
+			break;
+		else if (written < 0)
 			return xerror("vmsplice");
 		fprintf(stderr, "written len = %d\n", written);
 
-		len -= written;
+		v.iov_len -= written;
+		v.iov_base += written;
 	}
 
 	return 0;
@@ -98,7 +101,7 @@
 	unsigned char *buffer;
 	struct stat sb;
 	long page_size;
-	int i, ret;
+	int i;
 
 	if (fstat(STDOUT_FILENO, &sb) < 0)
 		return xerror("stat");
@@ -112,7 +115,7 @@
 		return xerror("_SC_PAGESIZE");
 
 	fprintf(stderr, "getpagesize = %d\n", getpagesize());
-	fprintf(stderr, "page size: %d bytes\n", page_size);
+	fprintf(stderr, "page size: %d bytes\n", (int) page_size);
 
 	buffer = malloc(2 * 65536);
 	buffer[0]='A';

-- 
Jens Axboe

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/