Jörn
-- With a PC, I always felt limited by the software available. On Unix, I am limited only by my knowledge. -- Peter J. Schoenster--- linux-2.4.20/fs/jffs2/compr_zlib.c~zlib_users 2002-11-29 00:53:15.000000000 +0100 +++ linux-2.4.20/fs/jffs2/compr_zlib.c 2003-05-30 16:16:26.000000000 +0200 @@ -56,33 +56,13 @@ */ #define STREAM_END_SPACE 12 -static DECLARE_MUTEX(deflate_sem); -static DECLARE_MUTEX(inflate_sem); -static void *deflate_workspace; -static void *inflate_workspace; - int __init jffs2_zlib_init(void) { - deflate_workspace = vmalloc(zlib_deflate_workspacesize()); - if (!deflate_workspace) { - printk(KERN_WARNING "Failed to allocate %d bytes for deflate workspace\n", zlib_deflate_workspacesize()); - return -ENOMEM; - } - D1(printk(KERN_DEBUG "Allocated %d bytes for deflate workspace\n", zlib_deflate_workspacesize())); - inflate_workspace = vmalloc(zlib_inflate_workspacesize()); - if (!inflate_workspace) { - printk(KERN_WARNING "Failed to allocate %d bytes for inflate workspace\n", zlib_inflate_workspacesize()); - vfree(deflate_workspace); - return -ENOMEM; - } - D1(printk(KERN_DEBUG "Allocated %d bytes for inflate workspace\n", zlib_inflate_workspacesize())); return 0; } void jffs2_zlib_exit(void) { - vfree(deflate_workspace); - vfree(inflate_workspace); } int zlib_compress(unsigned char *data_in, unsigned char *cpage_out, @@ -94,12 +74,10 @@ if (*dstlen <= STREAM_END_SPACE) return -1; - down(&deflate_sem); - strm.workspace = deflate_workspace; + strm.workspace = NULL; if (Z_OK != zlib_deflateInit(&strm, 3)) { printk(KERN_WARNING "deflateInit failed\n"); - up(&deflate_sem); return -1; } @@ -120,7 +98,6 @@ if (ret != Z_OK) { D1(printk(KERN_DEBUG "deflate in loop returned %d\n", ret)); zlib_deflateEnd(&strm); - up(&deflate_sem); return -1; } } @@ -128,7 +105,6 @@ strm.avail_in = 0; ret = zlib_deflate(&strm, Z_FINISH); zlib_deflateEnd(&strm); - up(&deflate_sem); if (ret != Z_STREAM_END) { D1(printk(KERN_DEBUG "final deflate returned %d\n", ret)); return -1; @@ -151,12 +127,10 @@ z_stream strm; int ret; - down(&inflate_sem); - strm.workspace = inflate_workspace; + strm.workspace = NULL; if (Z_OK != zlib_inflateInit(&strm)) { printk(KERN_WARNING "inflateInit failed\n"); - up(&inflate_sem); return; } strm.next_in = data_in; @@ -173,5 +147,4 @@ printk(KERN_NOTICE "inflate returned %d\n", ret); } zlib_inflateEnd(&strm); - up(&inflate_sem); } --- linux-2.4.20/fs/cramfs/uncompress.c~zlib_users 2002-11-29 00:53:15.000000000 +0100 +++ linux-2.4.20/fs/cramfs/uncompress.c 2003-05-30 16:17:22.000000000 +0200 @@ -34,19 +34,21 @@ stream.next_out = dst; stream.avail_out = dstlen; - err = zlib_inflateReset(&stream); - if (err != Z_OK) { - printk("zlib_inflateReset error %d\n", err); - zlib_inflateEnd(&stream); - zlib_inflateInit(&stream); - } + stream.workspace = NULL; + stream.next_in = NULL; + stream.avail_in = 0; + err = zlib_inflateInit(&stream); + if (err != Z_OK) + goto err; err = zlib_inflate(&stream, Z_FINISH); if (err != Z_STREAM_END) goto err; + zlib_inflateEnd(&stream); return stream.total_out; err: + zlib_inflateEnd(&stream); printk("Error %d while decompressing!\n", err); printk("%p(%d)->%p(%d)\n", src, srclen, dst, dstlen); return 0; @@ -54,24 +56,10 @@ int cramfs_uncompress_init(void) { - if (!initialized++) { - stream.workspace = vmalloc(zlib_inflate_workspacesize()); - if ( !stream.workspace ) { - initialized = 0; - return -ENOMEM; - } - stream.next_in = NULL; - stream.avail_in = 0; - zlib_inflateInit(&stream); - } return 0; } int cramfs_uncompress_exit(void) { - if (!--initialized) { - zlib_inflateEnd(&stream); - vfree(stream.workspace); - } return 0; } --- linux-2.4.20/fs/isofs/compress.c~zlib_users 2003-05-27 16:51:45.000000000 +0200 +++ linux-2.4.20/fs/isofs/compress.c 2003-05-30 16:16:26.000000000 +0200 @@ -48,13 +48,6 @@ static char zisofs_sink_page[PAGE_CACHE_SIZE]; /* - * This contains the zlib memory allocation and the mutex for the - * allocation; this avoids failures at block-decompression time. - */ -static void *zisofs_zlib_workspace; -static struct semaphore zisofs_zlib_semaphore; - -/* * When decompressing, we typically obtain more than one page * per reference. We inject the additional pages into the page * cache as a form of readahead. @@ -206,8 +199,7 @@ stream.avail_in = min(bufsize-(cstart & bufmask), csize); csize -= stream.avail_in; - stream.workspace = zisofs_zlib_workspace; - down(&zisofs_zlib_semaphore); + stream.workspace = NULL; zerr = zlib_inflateInit(&stream); if ( zerr != Z_OK ) { @@ -215,7 +207,7 @@ err = -ENOMEM; printk(KERN_DEBUG "zisofs: zisofs_inflateInit returned %d\n", zerr); - goto z_eio; + goto b_eio; } while ( !bail && fpage < maxpage ) { @@ -293,9 +285,6 @@ } zlib_inflateEnd(&stream); - z_eio: - up(&zisofs_zlib_semaphore); - b_eio: for ( i = 0 ; i < haveblocks ; i++ ) { if ( bhs[i] ) @@ -339,11 +328,6 @@ return 0; } - zisofs_zlib_workspace = vmalloc(zlib_inflate_workspacesize()); - if ( !zisofs_zlib_workspace ) - return -ENOMEM; - init_MUTEX(&zisofs_zlib_semaphore); - initialized = 1; return 0; } @@ -355,6 +339,5 @@ return; } - vfree(zisofs_zlib_workspace); initialized = 0; } - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/