u-boot/lib/lz4_wrapper.c

110 lines
2.6 KiB
C
Raw Normal View History

// SPDX-License-Identifier: GPL 2.0+ OR BSD-3-Clause
/*
* Copyright 2015 Google Inc.
*/
#include <common.h>
#include <compiler.h>
#include <image.h>
#include <linux/kernel.h>
#include <linux/types.h>
lz4: fix decompressor on big-endian powerpc Booting an lz4-compressed kernel image fails on our powerpc board with -EPROTONOSUPPORT. Adding a bit of debug prints, we get magic: 0x184d2204 flags: 0x64 reserved0: 1 has_content_checksum: 1 has_content_size: 0 has_block_checksum: 0 independent_blocks: 1 version: 0 block_descriptor: 70 reserved1: 7 max_block_size: 0 reserved2: 0 So the magic is ok, but the version check fails, also some reserved bits are apparently set. But that's because the code interprets the "flags" and "block_descriptor" bytes wrongly: Using bit-fields to access individual bits of an "on the wire" format is not portable, not even when restricted to the C flavour implemented by gcc. Quoting the gcc manual: * 'The order of allocation of bit-fields within a unit (C90 6.5.2.1, C99 and C11 6.7.2.1).' Determined by ABI. and indeed, the PPC Processor ABI supplement says * Bit-fields are allocated from right to left (least to most significant) on Little-Endian implementations and from left to right (most to least significant) on Big-Endian implementations. The upstream code (github.com/lz4/lz4) uses explicit shifts and masks for encoding/decoding: /* FLG Byte */ *dstPtr++ = (BYTE)(((1 & _2BITS) << 6) /* Version('01') */ + ((cctxPtr->prefs.frameInfo.blockMode & _1BIT ) << 5) + ((cctxPtr->prefs.frameInfo.blockChecksumFlag & _1BIT ) << 4) + ((unsigned)(cctxPtr->prefs.frameInfo.contentSize > 0) << 3) + ((cctxPtr->prefs.frameInfo.contentChecksumFlag & _1BIT ) << 2) + (cctxPtr->prefs.frameInfo.dictID > 0) ); /* Flags */ { U32 const FLG = srcPtr[4]; U32 const version = (FLG>>6) & _2BITS; blockChecksumFlag = (FLG>>4) & _1BIT; blockMode = (FLG>>5) & _1BIT; contentSizeFlag = (FLG>>3) & _1BIT; contentChecksumFlag = (FLG>>2) & _1BIT; dictIDFlag = FLG & _1BIT; /* validate */ if (((FLG>>1)&_1BIT) != 0) return err0r(LZ4F_ERROR_reservedFlag_set); /* Reserved bit */ if (version != 1) return err0r(LZ4F_ERROR_headerVersion_wrong); /* Version Number, only supported value */ } Do the same here, and while at it, be more careful to use unaligned accessors to what is most likely unaligned. Also update the comment to make it clear that it only refers to the lz4.c file, not the following code of lz4_wrapper.c. This has been tested partly, of course, by seeing that my lz4-compressed kernel now boots, partly by running the (de)compression test-suite in the (x86_64) sandbox - i.e., it should still work just fine on little-endian hosts. Reviewed-by: Julius Werner <jwerner@chromium.org> Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
2020-06-07 20:29:18 +08:00
#include <asm/unaligned.h>
#include <u-boot/lz4.h>
lz4: fix decompressor on big-endian powerpc Booting an lz4-compressed kernel image fails on our powerpc board with -EPROTONOSUPPORT. Adding a bit of debug prints, we get magic: 0x184d2204 flags: 0x64 reserved0: 1 has_content_checksum: 1 has_content_size: 0 has_block_checksum: 0 independent_blocks: 1 version: 0 block_descriptor: 70 reserved1: 7 max_block_size: 0 reserved2: 0 So the magic is ok, but the version check fails, also some reserved bits are apparently set. But that's because the code interprets the "flags" and "block_descriptor" bytes wrongly: Using bit-fields to access individual bits of an "on the wire" format is not portable, not even when restricted to the C flavour implemented by gcc. Quoting the gcc manual: * 'The order of allocation of bit-fields within a unit (C90 6.5.2.1, C99 and C11 6.7.2.1).' Determined by ABI. and indeed, the PPC Processor ABI supplement says * Bit-fields are allocated from right to left (least to most significant) on Little-Endian implementations and from left to right (most to least significant) on Big-Endian implementations. The upstream code (github.com/lz4/lz4) uses explicit shifts and masks for encoding/decoding: /* FLG Byte */ *dstPtr++ = (BYTE)(((1 & _2BITS) << 6) /* Version('01') */ + ((cctxPtr->prefs.frameInfo.blockMode & _1BIT ) << 5) + ((cctxPtr->prefs.frameInfo.blockChecksumFlag & _1BIT ) << 4) + ((unsigned)(cctxPtr->prefs.frameInfo.contentSize > 0) << 3) + ((cctxPtr->prefs.frameInfo.contentChecksumFlag & _1BIT ) << 2) + (cctxPtr->prefs.frameInfo.dictID > 0) ); /* Flags */ { U32 const FLG = srcPtr[4]; U32 const version = (FLG>>6) & _2BITS; blockChecksumFlag = (FLG>>4) & _1BIT; blockMode = (FLG>>5) & _1BIT; contentSizeFlag = (FLG>>3) & _1BIT; contentChecksumFlag = (FLG>>2) & _1BIT; dictIDFlag = FLG & _1BIT; /* validate */ if (((FLG>>1)&_1BIT) != 0) return err0r(LZ4F_ERROR_reservedFlag_set); /* Reserved bit */ if (version != 1) return err0r(LZ4F_ERROR_headerVersion_wrong); /* Version Number, only supported value */ } Do the same here, and while at it, be more careful to use unaligned accessors to what is most likely unaligned. Also update the comment to make it clear that it only refers to the lz4.c file, not the following code of lz4_wrapper.c. This has been tested partly, of course, by seeing that my lz4-compressed kernel now boots, partly by running the (de)compression test-suite in the (x86_64) sandbox - i.e., it should still work just fine on little-endian hosts. Reviewed-by: Julius Werner <jwerner@chromium.org> Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
2020-06-07 20:29:18 +08:00
/* lz4.c is unaltered (except removing unrelated code) from github.com/Cyan4973/lz4. */
#include "lz4.c" /* #include for inlining, do not link! */
lz4: fix decompressor on big-endian powerpc Booting an lz4-compressed kernel image fails on our powerpc board with -EPROTONOSUPPORT. Adding a bit of debug prints, we get magic: 0x184d2204 flags: 0x64 reserved0: 1 has_content_checksum: 1 has_content_size: 0 has_block_checksum: 0 independent_blocks: 1 version: 0 block_descriptor: 70 reserved1: 7 max_block_size: 0 reserved2: 0 So the magic is ok, but the version check fails, also some reserved bits are apparently set. But that's because the code interprets the "flags" and "block_descriptor" bytes wrongly: Using bit-fields to access individual bits of an "on the wire" format is not portable, not even when restricted to the C flavour implemented by gcc. Quoting the gcc manual: * 'The order of allocation of bit-fields within a unit (C90 6.5.2.1, C99 and C11 6.7.2.1).' Determined by ABI. and indeed, the PPC Processor ABI supplement says * Bit-fields are allocated from right to left (least to most significant) on Little-Endian implementations and from left to right (most to least significant) on Big-Endian implementations. The upstream code (github.com/lz4/lz4) uses explicit shifts and masks for encoding/decoding: /* FLG Byte */ *dstPtr++ = (BYTE)(((1 & _2BITS) << 6) /* Version('01') */ + ((cctxPtr->prefs.frameInfo.blockMode & _1BIT ) << 5) + ((cctxPtr->prefs.frameInfo.blockChecksumFlag & _1BIT ) << 4) + ((unsigned)(cctxPtr->prefs.frameInfo.contentSize > 0) << 3) + ((cctxPtr->prefs.frameInfo.contentChecksumFlag & _1BIT ) << 2) + (cctxPtr->prefs.frameInfo.dictID > 0) ); /* Flags */ { U32 const FLG = srcPtr[4]; U32 const version = (FLG>>6) & _2BITS; blockChecksumFlag = (FLG>>4) & _1BIT; blockMode = (FLG>>5) & _1BIT; contentSizeFlag = (FLG>>3) & _1BIT; contentChecksumFlag = (FLG>>2) & _1BIT; dictIDFlag = FLG & _1BIT; /* validate */ if (((FLG>>1)&_1BIT) != 0) return err0r(LZ4F_ERROR_reservedFlag_set); /* Reserved bit */ if (version != 1) return err0r(LZ4F_ERROR_headerVersion_wrong); /* Version Number, only supported value */ } Do the same here, and while at it, be more careful to use unaligned accessors to what is most likely unaligned. Also update the comment to make it clear that it only refers to the lz4.c file, not the following code of lz4_wrapper.c. This has been tested partly, of course, by seeing that my lz4-compressed kernel now boots, partly by running the (de)compression test-suite in the (x86_64) sandbox - i.e., it should still work just fine on little-endian hosts. Reviewed-by: Julius Werner <jwerner@chromium.org> Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
2020-06-07 20:29:18 +08:00
#define LZ4F_BLOCKUNCOMPRESSED_FLAG 0x80000000U
int ulz4fn(const void *src, size_t srcn, void *dst, size_t *dstn)
{
const void *end = dst + *dstn;
const void *in = src;
void *out = dst;
int has_block_checksum;
int ret;
*dstn = 0;
{ /* With in-place decompression the header may become invalid later. */
lz4: fix decompressor on big-endian powerpc Booting an lz4-compressed kernel image fails on our powerpc board with -EPROTONOSUPPORT. Adding a bit of debug prints, we get magic: 0x184d2204 flags: 0x64 reserved0: 1 has_content_checksum: 1 has_content_size: 0 has_block_checksum: 0 independent_blocks: 1 version: 0 block_descriptor: 70 reserved1: 7 max_block_size: 0 reserved2: 0 So the magic is ok, but the version check fails, also some reserved bits are apparently set. But that's because the code interprets the "flags" and "block_descriptor" bytes wrongly: Using bit-fields to access individual bits of an "on the wire" format is not portable, not even when restricted to the C flavour implemented by gcc. Quoting the gcc manual: * 'The order of allocation of bit-fields within a unit (C90 6.5.2.1, C99 and C11 6.7.2.1).' Determined by ABI. and indeed, the PPC Processor ABI supplement says * Bit-fields are allocated from right to left (least to most significant) on Little-Endian implementations and from left to right (most to least significant) on Big-Endian implementations. The upstream code (github.com/lz4/lz4) uses explicit shifts and masks for encoding/decoding: /* FLG Byte */ *dstPtr++ = (BYTE)(((1 & _2BITS) << 6) /* Version('01') */ + ((cctxPtr->prefs.frameInfo.blockMode & _1BIT ) << 5) + ((cctxPtr->prefs.frameInfo.blockChecksumFlag & _1BIT ) << 4) + ((unsigned)(cctxPtr->prefs.frameInfo.contentSize > 0) << 3) + ((cctxPtr->prefs.frameInfo.contentChecksumFlag & _1BIT ) << 2) + (cctxPtr->prefs.frameInfo.dictID > 0) ); /* Flags */ { U32 const FLG = srcPtr[4]; U32 const version = (FLG>>6) & _2BITS; blockChecksumFlag = (FLG>>4) & _1BIT; blockMode = (FLG>>5) & _1BIT; contentSizeFlag = (FLG>>3) & _1BIT; contentChecksumFlag = (FLG>>2) & _1BIT; dictIDFlag = FLG & _1BIT; /* validate */ if (((FLG>>1)&_1BIT) != 0) return err0r(LZ4F_ERROR_reservedFlag_set); /* Reserved bit */ if (version != 1) return err0r(LZ4F_ERROR_headerVersion_wrong); /* Version Number, only supported value */ } Do the same here, and while at it, be more careful to use unaligned accessors to what is most likely unaligned. Also update the comment to make it clear that it only refers to the lz4.c file, not the following code of lz4_wrapper.c. This has been tested partly, of course, by seeing that my lz4-compressed kernel now boots, partly by running the (de)compression test-suite in the (x86_64) sandbox - i.e., it should still work just fine on little-endian hosts. Reviewed-by: Julius Werner <jwerner@chromium.org> Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
2020-06-07 20:29:18 +08:00
u32 magic;
u8 flags, version, independent_blocks, has_content_size;
u8 block_desc;
lz4: fix decompressor on big-endian powerpc Booting an lz4-compressed kernel image fails on our powerpc board with -EPROTONOSUPPORT. Adding a bit of debug prints, we get magic: 0x184d2204 flags: 0x64 reserved0: 1 has_content_checksum: 1 has_content_size: 0 has_block_checksum: 0 independent_blocks: 1 version: 0 block_descriptor: 70 reserved1: 7 max_block_size: 0 reserved2: 0 So the magic is ok, but the version check fails, also some reserved bits are apparently set. But that's because the code interprets the "flags" and "block_descriptor" bytes wrongly: Using bit-fields to access individual bits of an "on the wire" format is not portable, not even when restricted to the C flavour implemented by gcc. Quoting the gcc manual: * 'The order of allocation of bit-fields within a unit (C90 6.5.2.1, C99 and C11 6.7.2.1).' Determined by ABI. and indeed, the PPC Processor ABI supplement says * Bit-fields are allocated from right to left (least to most significant) on Little-Endian implementations and from left to right (most to least significant) on Big-Endian implementations. The upstream code (github.com/lz4/lz4) uses explicit shifts and masks for encoding/decoding: /* FLG Byte */ *dstPtr++ = (BYTE)(((1 & _2BITS) << 6) /* Version('01') */ + ((cctxPtr->prefs.frameInfo.blockMode & _1BIT ) << 5) + ((cctxPtr->prefs.frameInfo.blockChecksumFlag & _1BIT ) << 4) + ((unsigned)(cctxPtr->prefs.frameInfo.contentSize > 0) << 3) + ((cctxPtr->prefs.frameInfo.contentChecksumFlag & _1BIT ) << 2) + (cctxPtr->prefs.frameInfo.dictID > 0) ); /* Flags */ { U32 const FLG = srcPtr[4]; U32 const version = (FLG>>6) & _2BITS; blockChecksumFlag = (FLG>>4) & _1BIT; blockMode = (FLG>>5) & _1BIT; contentSizeFlag = (FLG>>3) & _1BIT; contentChecksumFlag = (FLG>>2) & _1BIT; dictIDFlag = FLG & _1BIT; /* validate */ if (((FLG>>1)&_1BIT) != 0) return err0r(LZ4F_ERROR_reservedFlag_set); /* Reserved bit */ if (version != 1) return err0r(LZ4F_ERROR_headerVersion_wrong); /* Version Number, only supported value */ } Do the same here, and while at it, be more careful to use unaligned accessors to what is most likely unaligned. Also update the comment to make it clear that it only refers to the lz4.c file, not the following code of lz4_wrapper.c. This has been tested partly, of course, by seeing that my lz4-compressed kernel now boots, partly by running the (de)compression test-suite in the (x86_64) sandbox - i.e., it should still work just fine on little-endian hosts. Reviewed-by: Julius Werner <jwerner@chromium.org> Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
2020-06-07 20:29:18 +08:00
if (srcn < sizeof(u32) + 3*sizeof(u8))
return -EINVAL; /* input overrun */
lz4: fix decompressor on big-endian powerpc Booting an lz4-compressed kernel image fails on our powerpc board with -EPROTONOSUPPORT. Adding a bit of debug prints, we get magic: 0x184d2204 flags: 0x64 reserved0: 1 has_content_checksum: 1 has_content_size: 0 has_block_checksum: 0 independent_blocks: 1 version: 0 block_descriptor: 70 reserved1: 7 max_block_size: 0 reserved2: 0 So the magic is ok, but the version check fails, also some reserved bits are apparently set. But that's because the code interprets the "flags" and "block_descriptor" bytes wrongly: Using bit-fields to access individual bits of an "on the wire" format is not portable, not even when restricted to the C flavour implemented by gcc. Quoting the gcc manual: * 'The order of allocation of bit-fields within a unit (C90 6.5.2.1, C99 and C11 6.7.2.1).' Determined by ABI. and indeed, the PPC Processor ABI supplement says * Bit-fields are allocated from right to left (least to most significant) on Little-Endian implementations and from left to right (most to least significant) on Big-Endian implementations. The upstream code (github.com/lz4/lz4) uses explicit shifts and masks for encoding/decoding: /* FLG Byte */ *dstPtr++ = (BYTE)(((1 & _2BITS) << 6) /* Version('01') */ + ((cctxPtr->prefs.frameInfo.blockMode & _1BIT ) << 5) + ((cctxPtr->prefs.frameInfo.blockChecksumFlag & _1BIT ) << 4) + ((unsigned)(cctxPtr->prefs.frameInfo.contentSize > 0) << 3) + ((cctxPtr->prefs.frameInfo.contentChecksumFlag & _1BIT ) << 2) + (cctxPtr->prefs.frameInfo.dictID > 0) ); /* Flags */ { U32 const FLG = srcPtr[4]; U32 const version = (FLG>>6) & _2BITS; blockChecksumFlag = (FLG>>4) & _1BIT; blockMode = (FLG>>5) & _1BIT; contentSizeFlag = (FLG>>3) & _1BIT; contentChecksumFlag = (FLG>>2) & _1BIT; dictIDFlag = FLG & _1BIT; /* validate */ if (((FLG>>1)&_1BIT) != 0) return err0r(LZ4F_ERROR_reservedFlag_set); /* Reserved bit */ if (version != 1) return err0r(LZ4F_ERROR_headerVersion_wrong); /* Version Number, only supported value */ } Do the same here, and while at it, be more careful to use unaligned accessors to what is most likely unaligned. Also update the comment to make it clear that it only refers to the lz4.c file, not the following code of lz4_wrapper.c. This has been tested partly, of course, by seeing that my lz4-compressed kernel now boots, partly by running the (de)compression test-suite in the (x86_64) sandbox - i.e., it should still work just fine on little-endian hosts. Reviewed-by: Julius Werner <jwerner@chromium.org> Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
2020-06-07 20:29:18 +08:00
magic = get_unaligned_le32(in);
in += sizeof(u32);
flags = *(u8 *)in;
in += sizeof(u8);
block_desc = *(u8 *)in;
in += sizeof(u8);
version = (flags >> 6) & 0x3;
independent_blocks = (flags >> 5) & 0x1;
has_block_checksum = (flags >> 4) & 0x1;
has_content_size = (flags >> 3) & 0x1;
/* We assume there's always only a single, standard frame. */
lz4: fix decompressor on big-endian powerpc Booting an lz4-compressed kernel image fails on our powerpc board with -EPROTONOSUPPORT. Adding a bit of debug prints, we get magic: 0x184d2204 flags: 0x64 reserved0: 1 has_content_checksum: 1 has_content_size: 0 has_block_checksum: 0 independent_blocks: 1 version: 0 block_descriptor: 70 reserved1: 7 max_block_size: 0 reserved2: 0 So the magic is ok, but the version check fails, also some reserved bits are apparently set. But that's because the code interprets the "flags" and "block_descriptor" bytes wrongly: Using bit-fields to access individual bits of an "on the wire" format is not portable, not even when restricted to the C flavour implemented by gcc. Quoting the gcc manual: * 'The order of allocation of bit-fields within a unit (C90 6.5.2.1, C99 and C11 6.7.2.1).' Determined by ABI. and indeed, the PPC Processor ABI supplement says * Bit-fields are allocated from right to left (least to most significant) on Little-Endian implementations and from left to right (most to least significant) on Big-Endian implementations. The upstream code (github.com/lz4/lz4) uses explicit shifts and masks for encoding/decoding: /* FLG Byte */ *dstPtr++ = (BYTE)(((1 & _2BITS) << 6) /* Version('01') */ + ((cctxPtr->prefs.frameInfo.blockMode & _1BIT ) << 5) + ((cctxPtr->prefs.frameInfo.blockChecksumFlag & _1BIT ) << 4) + ((unsigned)(cctxPtr->prefs.frameInfo.contentSize > 0) << 3) + ((cctxPtr->prefs.frameInfo.contentChecksumFlag & _1BIT ) << 2) + (cctxPtr->prefs.frameInfo.dictID > 0) ); /* Flags */ { U32 const FLG = srcPtr[4]; U32 const version = (FLG>>6) & _2BITS; blockChecksumFlag = (FLG>>4) & _1BIT; blockMode = (FLG>>5) & _1BIT; contentSizeFlag = (FLG>>3) & _1BIT; contentChecksumFlag = (FLG>>2) & _1BIT; dictIDFlag = FLG & _1BIT; /* validate */ if (((FLG>>1)&_1BIT) != 0) return err0r(LZ4F_ERROR_reservedFlag_set); /* Reserved bit */ if (version != 1) return err0r(LZ4F_ERROR_headerVersion_wrong); /* Version Number, only supported value */ } Do the same here, and while at it, be more careful to use unaligned accessors to what is most likely unaligned. Also update the comment to make it clear that it only refers to the lz4.c file, not the following code of lz4_wrapper.c. This has been tested partly, of course, by seeing that my lz4-compressed kernel now boots, partly by running the (de)compression test-suite in the (x86_64) sandbox - i.e., it should still work just fine on little-endian hosts. Reviewed-by: Julius Werner <jwerner@chromium.org> Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
2020-06-07 20:29:18 +08:00
if (magic != LZ4F_MAGIC || version != 1)
return -EPROTONOSUPPORT; /* unknown format */
lz4: fix decompressor on big-endian powerpc Booting an lz4-compressed kernel image fails on our powerpc board with -EPROTONOSUPPORT. Adding a bit of debug prints, we get magic: 0x184d2204 flags: 0x64 reserved0: 1 has_content_checksum: 1 has_content_size: 0 has_block_checksum: 0 independent_blocks: 1 version: 0 block_descriptor: 70 reserved1: 7 max_block_size: 0 reserved2: 0 So the magic is ok, but the version check fails, also some reserved bits are apparently set. But that's because the code interprets the "flags" and "block_descriptor" bytes wrongly: Using bit-fields to access individual bits of an "on the wire" format is not portable, not even when restricted to the C flavour implemented by gcc. Quoting the gcc manual: * 'The order of allocation of bit-fields within a unit (C90 6.5.2.1, C99 and C11 6.7.2.1).' Determined by ABI. and indeed, the PPC Processor ABI supplement says * Bit-fields are allocated from right to left (least to most significant) on Little-Endian implementations and from left to right (most to least significant) on Big-Endian implementations. The upstream code (github.com/lz4/lz4) uses explicit shifts and masks for encoding/decoding: /* FLG Byte */ *dstPtr++ = (BYTE)(((1 & _2BITS) << 6) /* Version('01') */ + ((cctxPtr->prefs.frameInfo.blockMode & _1BIT ) << 5) + ((cctxPtr->prefs.frameInfo.blockChecksumFlag & _1BIT ) << 4) + ((unsigned)(cctxPtr->prefs.frameInfo.contentSize > 0) << 3) + ((cctxPtr->prefs.frameInfo.contentChecksumFlag & _1BIT ) << 2) + (cctxPtr->prefs.frameInfo.dictID > 0) ); /* Flags */ { U32 const FLG = srcPtr[4]; U32 const version = (FLG>>6) & _2BITS; blockChecksumFlag = (FLG>>4) & _1BIT; blockMode = (FLG>>5) & _1BIT; contentSizeFlag = (FLG>>3) & _1BIT; contentChecksumFlag = (FLG>>2) & _1BIT; dictIDFlag = FLG & _1BIT; /* validate */ if (((FLG>>1)&_1BIT) != 0) return err0r(LZ4F_ERROR_reservedFlag_set); /* Reserved bit */ if (version != 1) return err0r(LZ4F_ERROR_headerVersion_wrong); /* Version Number, only supported value */ } Do the same here, and while at it, be more careful to use unaligned accessors to what is most likely unaligned. Also update the comment to make it clear that it only refers to the lz4.c file, not the following code of lz4_wrapper.c. This has been tested partly, of course, by seeing that my lz4-compressed kernel now boots, partly by running the (de)compression test-suite in the (x86_64) sandbox - i.e., it should still work just fine on little-endian hosts. Reviewed-by: Julius Werner <jwerner@chromium.org> Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
2020-06-07 20:29:18 +08:00
if ((flags & 0x03) || (block_desc & 0x8f))
return -EINVAL; /* reserved bits must be zero */
if (!independent_blocks)
return -EPROTONOSUPPORT; /* we can't support this yet */
lz4: fix decompressor on big-endian powerpc Booting an lz4-compressed kernel image fails on our powerpc board with -EPROTONOSUPPORT. Adding a bit of debug prints, we get magic: 0x184d2204 flags: 0x64 reserved0: 1 has_content_checksum: 1 has_content_size: 0 has_block_checksum: 0 independent_blocks: 1 version: 0 block_descriptor: 70 reserved1: 7 max_block_size: 0 reserved2: 0 So the magic is ok, but the version check fails, also some reserved bits are apparently set. But that's because the code interprets the "flags" and "block_descriptor" bytes wrongly: Using bit-fields to access individual bits of an "on the wire" format is not portable, not even when restricted to the C flavour implemented by gcc. Quoting the gcc manual: * 'The order of allocation of bit-fields within a unit (C90 6.5.2.1, C99 and C11 6.7.2.1).' Determined by ABI. and indeed, the PPC Processor ABI supplement says * Bit-fields are allocated from right to left (least to most significant) on Little-Endian implementations and from left to right (most to least significant) on Big-Endian implementations. The upstream code (github.com/lz4/lz4) uses explicit shifts and masks for encoding/decoding: /* FLG Byte */ *dstPtr++ = (BYTE)(((1 & _2BITS) << 6) /* Version('01') */ + ((cctxPtr->prefs.frameInfo.blockMode & _1BIT ) << 5) + ((cctxPtr->prefs.frameInfo.blockChecksumFlag & _1BIT ) << 4) + ((unsigned)(cctxPtr->prefs.frameInfo.contentSize > 0) << 3) + ((cctxPtr->prefs.frameInfo.contentChecksumFlag & _1BIT ) << 2) + (cctxPtr->prefs.frameInfo.dictID > 0) ); /* Flags */ { U32 const FLG = srcPtr[4]; U32 const version = (FLG>>6) & _2BITS; blockChecksumFlag = (FLG>>4) & _1BIT; blockMode = (FLG>>5) & _1BIT; contentSizeFlag = (FLG>>3) & _1BIT; contentChecksumFlag = (FLG>>2) & _1BIT; dictIDFlag = FLG & _1BIT; /* validate */ if (((FLG>>1)&_1BIT) != 0) return err0r(LZ4F_ERROR_reservedFlag_set); /* Reserved bit */ if (version != 1) return err0r(LZ4F_ERROR_headerVersion_wrong); /* Version Number, only supported value */ } Do the same here, and while at it, be more careful to use unaligned accessors to what is most likely unaligned. Also update the comment to make it clear that it only refers to the lz4.c file, not the following code of lz4_wrapper.c. This has been tested partly, of course, by seeing that my lz4-compressed kernel now boots, partly by running the (de)compression test-suite in the (x86_64) sandbox - i.e., it should still work just fine on little-endian hosts. Reviewed-by: Julius Werner <jwerner@chromium.org> Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
2020-06-07 20:29:18 +08:00
if (has_content_size) {
if (srcn < sizeof(u32) + 3*sizeof(u8) + sizeof(u64))
return -EINVAL; /* input overrun */
in += sizeof(u64);
lz4: fix decompressor on big-endian powerpc Booting an lz4-compressed kernel image fails on our powerpc board with -EPROTONOSUPPORT. Adding a bit of debug prints, we get magic: 0x184d2204 flags: 0x64 reserved0: 1 has_content_checksum: 1 has_content_size: 0 has_block_checksum: 0 independent_blocks: 1 version: 0 block_descriptor: 70 reserved1: 7 max_block_size: 0 reserved2: 0 So the magic is ok, but the version check fails, also some reserved bits are apparently set. But that's because the code interprets the "flags" and "block_descriptor" bytes wrongly: Using bit-fields to access individual bits of an "on the wire" format is not portable, not even when restricted to the C flavour implemented by gcc. Quoting the gcc manual: * 'The order of allocation of bit-fields within a unit (C90 6.5.2.1, C99 and C11 6.7.2.1).' Determined by ABI. and indeed, the PPC Processor ABI supplement says * Bit-fields are allocated from right to left (least to most significant) on Little-Endian implementations and from left to right (most to least significant) on Big-Endian implementations. The upstream code (github.com/lz4/lz4) uses explicit shifts and masks for encoding/decoding: /* FLG Byte */ *dstPtr++ = (BYTE)(((1 & _2BITS) << 6) /* Version('01') */ + ((cctxPtr->prefs.frameInfo.blockMode & _1BIT ) << 5) + ((cctxPtr->prefs.frameInfo.blockChecksumFlag & _1BIT ) << 4) + ((unsigned)(cctxPtr->prefs.frameInfo.contentSize > 0) << 3) + ((cctxPtr->prefs.frameInfo.contentChecksumFlag & _1BIT ) << 2) + (cctxPtr->prefs.frameInfo.dictID > 0) ); /* Flags */ { U32 const FLG = srcPtr[4]; U32 const version = (FLG>>6) & _2BITS; blockChecksumFlag = (FLG>>4) & _1BIT; blockMode = (FLG>>5) & _1BIT; contentSizeFlag = (FLG>>3) & _1BIT; contentChecksumFlag = (FLG>>2) & _1BIT; dictIDFlag = FLG & _1BIT; /* validate */ if (((FLG>>1)&_1BIT) != 0) return err0r(LZ4F_ERROR_reservedFlag_set); /* Reserved bit */ if (version != 1) return err0r(LZ4F_ERROR_headerVersion_wrong); /* Version Number, only supported value */ } Do the same here, and while at it, be more careful to use unaligned accessors to what is most likely unaligned. Also update the comment to make it clear that it only refers to the lz4.c file, not the following code of lz4_wrapper.c. This has been tested partly, of course, by seeing that my lz4-compressed kernel now boots, partly by running the (de)compression test-suite in the (x86_64) sandbox - i.e., it should still work just fine on little-endian hosts. Reviewed-by: Julius Werner <jwerner@chromium.org> Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
2020-06-07 20:29:18 +08:00
}
/* Header checksum byte */
in += sizeof(u8);
}
while (1) {
lz4: fix decompressor on big-endian powerpc Booting an lz4-compressed kernel image fails on our powerpc board with -EPROTONOSUPPORT. Adding a bit of debug prints, we get magic: 0x184d2204 flags: 0x64 reserved0: 1 has_content_checksum: 1 has_content_size: 0 has_block_checksum: 0 independent_blocks: 1 version: 0 block_descriptor: 70 reserved1: 7 max_block_size: 0 reserved2: 0 So the magic is ok, but the version check fails, also some reserved bits are apparently set. But that's because the code interprets the "flags" and "block_descriptor" bytes wrongly: Using bit-fields to access individual bits of an "on the wire" format is not portable, not even when restricted to the C flavour implemented by gcc. Quoting the gcc manual: * 'The order of allocation of bit-fields within a unit (C90 6.5.2.1, C99 and C11 6.7.2.1).' Determined by ABI. and indeed, the PPC Processor ABI supplement says * Bit-fields are allocated from right to left (least to most significant) on Little-Endian implementations and from left to right (most to least significant) on Big-Endian implementations. The upstream code (github.com/lz4/lz4) uses explicit shifts and masks for encoding/decoding: /* FLG Byte */ *dstPtr++ = (BYTE)(((1 & _2BITS) << 6) /* Version('01') */ + ((cctxPtr->prefs.frameInfo.blockMode & _1BIT ) << 5) + ((cctxPtr->prefs.frameInfo.blockChecksumFlag & _1BIT ) << 4) + ((unsigned)(cctxPtr->prefs.frameInfo.contentSize > 0) << 3) + ((cctxPtr->prefs.frameInfo.contentChecksumFlag & _1BIT ) << 2) + (cctxPtr->prefs.frameInfo.dictID > 0) ); /* Flags */ { U32 const FLG = srcPtr[4]; U32 const version = (FLG>>6) & _2BITS; blockChecksumFlag = (FLG>>4) & _1BIT; blockMode = (FLG>>5) & _1BIT; contentSizeFlag = (FLG>>3) & _1BIT; contentChecksumFlag = (FLG>>2) & _1BIT; dictIDFlag = FLG & _1BIT; /* validate */ if (((FLG>>1)&_1BIT) != 0) return err0r(LZ4F_ERROR_reservedFlag_set); /* Reserved bit */ if (version != 1) return err0r(LZ4F_ERROR_headerVersion_wrong); /* Version Number, only supported value */ } Do the same here, and while at it, be more careful to use unaligned accessors to what is most likely unaligned. Also update the comment to make it clear that it only refers to the lz4.c file, not the following code of lz4_wrapper.c. This has been tested partly, of course, by seeing that my lz4-compressed kernel now boots, partly by running the (de)compression test-suite in the (x86_64) sandbox - i.e., it should still work just fine on little-endian hosts. Reviewed-by: Julius Werner <jwerner@chromium.org> Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
2020-06-07 20:29:18 +08:00
u32 block_header, block_size;
lz4: fix decompressor on big-endian powerpc Booting an lz4-compressed kernel image fails on our powerpc board with -EPROTONOSUPPORT. Adding a bit of debug prints, we get magic: 0x184d2204 flags: 0x64 reserved0: 1 has_content_checksum: 1 has_content_size: 0 has_block_checksum: 0 independent_blocks: 1 version: 0 block_descriptor: 70 reserved1: 7 max_block_size: 0 reserved2: 0 So the magic is ok, but the version check fails, also some reserved bits are apparently set. But that's because the code interprets the "flags" and "block_descriptor" bytes wrongly: Using bit-fields to access individual bits of an "on the wire" format is not portable, not even when restricted to the C flavour implemented by gcc. Quoting the gcc manual: * 'The order of allocation of bit-fields within a unit (C90 6.5.2.1, C99 and C11 6.7.2.1).' Determined by ABI. and indeed, the PPC Processor ABI supplement says * Bit-fields are allocated from right to left (least to most significant) on Little-Endian implementations and from left to right (most to least significant) on Big-Endian implementations. The upstream code (github.com/lz4/lz4) uses explicit shifts and masks for encoding/decoding: /* FLG Byte */ *dstPtr++ = (BYTE)(((1 & _2BITS) << 6) /* Version('01') */ + ((cctxPtr->prefs.frameInfo.blockMode & _1BIT ) << 5) + ((cctxPtr->prefs.frameInfo.blockChecksumFlag & _1BIT ) << 4) + ((unsigned)(cctxPtr->prefs.frameInfo.contentSize > 0) << 3) + ((cctxPtr->prefs.frameInfo.contentChecksumFlag & _1BIT ) << 2) + (cctxPtr->prefs.frameInfo.dictID > 0) ); /* Flags */ { U32 const FLG = srcPtr[4]; U32 const version = (FLG>>6) & _2BITS; blockChecksumFlag = (FLG>>4) & _1BIT; blockMode = (FLG>>5) & _1BIT; contentSizeFlag = (FLG>>3) & _1BIT; contentChecksumFlag = (FLG>>2) & _1BIT; dictIDFlag = FLG & _1BIT; /* validate */ if (((FLG>>1)&_1BIT) != 0) return err0r(LZ4F_ERROR_reservedFlag_set); /* Reserved bit */ if (version != 1) return err0r(LZ4F_ERROR_headerVersion_wrong); /* Version Number, only supported value */ } Do the same here, and while at it, be more careful to use unaligned accessors to what is most likely unaligned. Also update the comment to make it clear that it only refers to the lz4.c file, not the following code of lz4_wrapper.c. This has been tested partly, of course, by seeing that my lz4-compressed kernel now boots, partly by running the (de)compression test-suite in the (x86_64) sandbox - i.e., it should still work just fine on little-endian hosts. Reviewed-by: Julius Werner <jwerner@chromium.org> Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
2020-06-07 20:29:18 +08:00
block_header = get_unaligned_le32(in);
in += sizeof(u32);
block_size = block_header & ~LZ4F_BLOCKUNCOMPRESSED_FLAG;
lz4: fix decompressor on big-endian powerpc Booting an lz4-compressed kernel image fails on our powerpc board with -EPROTONOSUPPORT. Adding a bit of debug prints, we get magic: 0x184d2204 flags: 0x64 reserved0: 1 has_content_checksum: 1 has_content_size: 0 has_block_checksum: 0 independent_blocks: 1 version: 0 block_descriptor: 70 reserved1: 7 max_block_size: 0 reserved2: 0 So the magic is ok, but the version check fails, also some reserved bits are apparently set. But that's because the code interprets the "flags" and "block_descriptor" bytes wrongly: Using bit-fields to access individual bits of an "on the wire" format is not portable, not even when restricted to the C flavour implemented by gcc. Quoting the gcc manual: * 'The order of allocation of bit-fields within a unit (C90 6.5.2.1, C99 and C11 6.7.2.1).' Determined by ABI. and indeed, the PPC Processor ABI supplement says * Bit-fields are allocated from right to left (least to most significant) on Little-Endian implementations and from left to right (most to least significant) on Big-Endian implementations. The upstream code (github.com/lz4/lz4) uses explicit shifts and masks for encoding/decoding: /* FLG Byte */ *dstPtr++ = (BYTE)(((1 & _2BITS) << 6) /* Version('01') */ + ((cctxPtr->prefs.frameInfo.blockMode & _1BIT ) << 5) + ((cctxPtr->prefs.frameInfo.blockChecksumFlag & _1BIT ) << 4) + ((unsigned)(cctxPtr->prefs.frameInfo.contentSize > 0) << 3) + ((cctxPtr->prefs.frameInfo.contentChecksumFlag & _1BIT ) << 2) + (cctxPtr->prefs.frameInfo.dictID > 0) ); /* Flags */ { U32 const FLG = srcPtr[4]; U32 const version = (FLG>>6) & _2BITS; blockChecksumFlag = (FLG>>4) & _1BIT; blockMode = (FLG>>5) & _1BIT; contentSizeFlag = (FLG>>3) & _1BIT; contentChecksumFlag = (FLG>>2) & _1BIT; dictIDFlag = FLG & _1BIT; /* validate */ if (((FLG>>1)&_1BIT) != 0) return err0r(LZ4F_ERROR_reservedFlag_set); /* Reserved bit */ if (version != 1) return err0r(LZ4F_ERROR_headerVersion_wrong); /* Version Number, only supported value */ } Do the same here, and while at it, be more careful to use unaligned accessors to what is most likely unaligned. Also update the comment to make it clear that it only refers to the lz4.c file, not the following code of lz4_wrapper.c. This has been tested partly, of course, by seeing that my lz4-compressed kernel now boots, partly by running the (de)compression test-suite in the (x86_64) sandbox - i.e., it should still work just fine on little-endian hosts. Reviewed-by: Julius Werner <jwerner@chromium.org> Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
2020-06-07 20:29:18 +08:00
if (in - src + block_size > srcn) {
ret = -EINVAL; /* input overrun */
break;
}
lz4: fix decompressor on big-endian powerpc Booting an lz4-compressed kernel image fails on our powerpc board with -EPROTONOSUPPORT. Adding a bit of debug prints, we get magic: 0x184d2204 flags: 0x64 reserved0: 1 has_content_checksum: 1 has_content_size: 0 has_block_checksum: 0 independent_blocks: 1 version: 0 block_descriptor: 70 reserved1: 7 max_block_size: 0 reserved2: 0 So the magic is ok, but the version check fails, also some reserved bits are apparently set. But that's because the code interprets the "flags" and "block_descriptor" bytes wrongly: Using bit-fields to access individual bits of an "on the wire" format is not portable, not even when restricted to the C flavour implemented by gcc. Quoting the gcc manual: * 'The order of allocation of bit-fields within a unit (C90 6.5.2.1, C99 and C11 6.7.2.1).' Determined by ABI. and indeed, the PPC Processor ABI supplement says * Bit-fields are allocated from right to left (least to most significant) on Little-Endian implementations and from left to right (most to least significant) on Big-Endian implementations. The upstream code (github.com/lz4/lz4) uses explicit shifts and masks for encoding/decoding: /* FLG Byte */ *dstPtr++ = (BYTE)(((1 & _2BITS) << 6) /* Version('01') */ + ((cctxPtr->prefs.frameInfo.blockMode & _1BIT ) << 5) + ((cctxPtr->prefs.frameInfo.blockChecksumFlag & _1BIT ) << 4) + ((unsigned)(cctxPtr->prefs.frameInfo.contentSize > 0) << 3) + ((cctxPtr->prefs.frameInfo.contentChecksumFlag & _1BIT ) << 2) + (cctxPtr->prefs.frameInfo.dictID > 0) ); /* Flags */ { U32 const FLG = srcPtr[4]; U32 const version = (FLG>>6) & _2BITS; blockChecksumFlag = (FLG>>4) & _1BIT; blockMode = (FLG>>5) & _1BIT; contentSizeFlag = (FLG>>3) & _1BIT; contentChecksumFlag = (FLG>>2) & _1BIT; dictIDFlag = FLG & _1BIT; /* validate */ if (((FLG>>1)&_1BIT) != 0) return err0r(LZ4F_ERROR_reservedFlag_set); /* Reserved bit */ if (version != 1) return err0r(LZ4F_ERROR_headerVersion_wrong); /* Version Number, only supported value */ } Do the same here, and while at it, be more careful to use unaligned accessors to what is most likely unaligned. Also update the comment to make it clear that it only refers to the lz4.c file, not the following code of lz4_wrapper.c. This has been tested partly, of course, by seeing that my lz4-compressed kernel now boots, partly by running the (de)compression test-suite in the (x86_64) sandbox - i.e., it should still work just fine on little-endian hosts. Reviewed-by: Julius Werner <jwerner@chromium.org> Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
2020-06-07 20:29:18 +08:00
if (!block_size) {
ret = 0; /* decompression successful */
break;
}
lz4: fix decompressor on big-endian powerpc Booting an lz4-compressed kernel image fails on our powerpc board with -EPROTONOSUPPORT. Adding a bit of debug prints, we get magic: 0x184d2204 flags: 0x64 reserved0: 1 has_content_checksum: 1 has_content_size: 0 has_block_checksum: 0 independent_blocks: 1 version: 0 block_descriptor: 70 reserved1: 7 max_block_size: 0 reserved2: 0 So the magic is ok, but the version check fails, also some reserved bits are apparently set. But that's because the code interprets the "flags" and "block_descriptor" bytes wrongly: Using bit-fields to access individual bits of an "on the wire" format is not portable, not even when restricted to the C flavour implemented by gcc. Quoting the gcc manual: * 'The order of allocation of bit-fields within a unit (C90 6.5.2.1, C99 and C11 6.7.2.1).' Determined by ABI. and indeed, the PPC Processor ABI supplement says * Bit-fields are allocated from right to left (least to most significant) on Little-Endian implementations and from left to right (most to least significant) on Big-Endian implementations. The upstream code (github.com/lz4/lz4) uses explicit shifts and masks for encoding/decoding: /* FLG Byte */ *dstPtr++ = (BYTE)(((1 & _2BITS) << 6) /* Version('01') */ + ((cctxPtr->prefs.frameInfo.blockMode & _1BIT ) << 5) + ((cctxPtr->prefs.frameInfo.blockChecksumFlag & _1BIT ) << 4) + ((unsigned)(cctxPtr->prefs.frameInfo.contentSize > 0) << 3) + ((cctxPtr->prefs.frameInfo.contentChecksumFlag & _1BIT ) << 2) + (cctxPtr->prefs.frameInfo.dictID > 0) ); /* Flags */ { U32 const FLG = srcPtr[4]; U32 const version = (FLG>>6) & _2BITS; blockChecksumFlag = (FLG>>4) & _1BIT; blockMode = (FLG>>5) & _1BIT; contentSizeFlag = (FLG>>3) & _1BIT; contentChecksumFlag = (FLG>>2) & _1BIT; dictIDFlag = FLG & _1BIT; /* validate */ if (((FLG>>1)&_1BIT) != 0) return err0r(LZ4F_ERROR_reservedFlag_set); /* Reserved bit */ if (version != 1) return err0r(LZ4F_ERROR_headerVersion_wrong); /* Version Number, only supported value */ } Do the same here, and while at it, be more careful to use unaligned accessors to what is most likely unaligned. Also update the comment to make it clear that it only refers to the lz4.c file, not the following code of lz4_wrapper.c. This has been tested partly, of course, by seeing that my lz4-compressed kernel now boots, partly by running the (de)compression test-suite in the (x86_64) sandbox - i.e., it should still work just fine on little-endian hosts. Reviewed-by: Julius Werner <jwerner@chromium.org> Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
2020-06-07 20:29:18 +08:00
if (block_header & LZ4F_BLOCKUNCOMPRESSED_FLAG) {
size_t size = min((ptrdiff_t)block_size, end - out);
memcpy(out, in, size);
out += size;
lz4: fix decompressor on big-endian powerpc Booting an lz4-compressed kernel image fails on our powerpc board with -EPROTONOSUPPORT. Adding a bit of debug prints, we get magic: 0x184d2204 flags: 0x64 reserved0: 1 has_content_checksum: 1 has_content_size: 0 has_block_checksum: 0 independent_blocks: 1 version: 0 block_descriptor: 70 reserved1: 7 max_block_size: 0 reserved2: 0 So the magic is ok, but the version check fails, also some reserved bits are apparently set. But that's because the code interprets the "flags" and "block_descriptor" bytes wrongly: Using bit-fields to access individual bits of an "on the wire" format is not portable, not even when restricted to the C flavour implemented by gcc. Quoting the gcc manual: * 'The order of allocation of bit-fields within a unit (C90 6.5.2.1, C99 and C11 6.7.2.1).' Determined by ABI. and indeed, the PPC Processor ABI supplement says * Bit-fields are allocated from right to left (least to most significant) on Little-Endian implementations and from left to right (most to least significant) on Big-Endian implementations. The upstream code (github.com/lz4/lz4) uses explicit shifts and masks for encoding/decoding: /* FLG Byte */ *dstPtr++ = (BYTE)(((1 & _2BITS) << 6) /* Version('01') */ + ((cctxPtr->prefs.frameInfo.blockMode & _1BIT ) << 5) + ((cctxPtr->prefs.frameInfo.blockChecksumFlag & _1BIT ) << 4) + ((unsigned)(cctxPtr->prefs.frameInfo.contentSize > 0) << 3) + ((cctxPtr->prefs.frameInfo.contentChecksumFlag & _1BIT ) << 2) + (cctxPtr->prefs.frameInfo.dictID > 0) ); /* Flags */ { U32 const FLG = srcPtr[4]; U32 const version = (FLG>>6) & _2BITS; blockChecksumFlag = (FLG>>4) & _1BIT; blockMode = (FLG>>5) & _1BIT; contentSizeFlag = (FLG>>3) & _1BIT; contentChecksumFlag = (FLG>>2) & _1BIT; dictIDFlag = FLG & _1BIT; /* validate */ if (((FLG>>1)&_1BIT) != 0) return err0r(LZ4F_ERROR_reservedFlag_set); /* Reserved bit */ if (version != 1) return err0r(LZ4F_ERROR_headerVersion_wrong); /* Version Number, only supported value */ } Do the same here, and while at it, be more careful to use unaligned accessors to what is most likely unaligned. Also update the comment to make it clear that it only refers to the lz4.c file, not the following code of lz4_wrapper.c. This has been tested partly, of course, by seeing that my lz4-compressed kernel now boots, partly by running the (de)compression test-suite in the (x86_64) sandbox - i.e., it should still work just fine on little-endian hosts. Reviewed-by: Julius Werner <jwerner@chromium.org> Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
2020-06-07 20:29:18 +08:00
if (size < block_size) {
ret = -ENOBUFS; /* output overrun */
break;
}
} else {
/* constant folding essential, do not touch params! */
lz4: fix decompressor on big-endian powerpc Booting an lz4-compressed kernel image fails on our powerpc board with -EPROTONOSUPPORT. Adding a bit of debug prints, we get magic: 0x184d2204 flags: 0x64 reserved0: 1 has_content_checksum: 1 has_content_size: 0 has_block_checksum: 0 independent_blocks: 1 version: 0 block_descriptor: 70 reserved1: 7 max_block_size: 0 reserved2: 0 So the magic is ok, but the version check fails, also some reserved bits are apparently set. But that's because the code interprets the "flags" and "block_descriptor" bytes wrongly: Using bit-fields to access individual bits of an "on the wire" format is not portable, not even when restricted to the C flavour implemented by gcc. Quoting the gcc manual: * 'The order of allocation of bit-fields within a unit (C90 6.5.2.1, C99 and C11 6.7.2.1).' Determined by ABI. and indeed, the PPC Processor ABI supplement says * Bit-fields are allocated from right to left (least to most significant) on Little-Endian implementations and from left to right (most to least significant) on Big-Endian implementations. The upstream code (github.com/lz4/lz4) uses explicit shifts and masks for encoding/decoding: /* FLG Byte */ *dstPtr++ = (BYTE)(((1 & _2BITS) << 6) /* Version('01') */ + ((cctxPtr->prefs.frameInfo.blockMode & _1BIT ) << 5) + ((cctxPtr->prefs.frameInfo.blockChecksumFlag & _1BIT ) << 4) + ((unsigned)(cctxPtr->prefs.frameInfo.contentSize > 0) << 3) + ((cctxPtr->prefs.frameInfo.contentChecksumFlag & _1BIT ) << 2) + (cctxPtr->prefs.frameInfo.dictID > 0) ); /* Flags */ { U32 const FLG = srcPtr[4]; U32 const version = (FLG>>6) & _2BITS; blockChecksumFlag = (FLG>>4) & _1BIT; blockMode = (FLG>>5) & _1BIT; contentSizeFlag = (FLG>>3) & _1BIT; contentChecksumFlag = (FLG>>2) & _1BIT; dictIDFlag = FLG & _1BIT; /* validate */ if (((FLG>>1)&_1BIT) != 0) return err0r(LZ4F_ERROR_reservedFlag_set); /* Reserved bit */ if (version != 1) return err0r(LZ4F_ERROR_headerVersion_wrong); /* Version Number, only supported value */ } Do the same here, and while at it, be more careful to use unaligned accessors to what is most likely unaligned. Also update the comment to make it clear that it only refers to the lz4.c file, not the following code of lz4_wrapper.c. This has been tested partly, of course, by seeing that my lz4-compressed kernel now boots, partly by running the (de)compression test-suite in the (x86_64) sandbox - i.e., it should still work just fine on little-endian hosts. Reviewed-by: Julius Werner <jwerner@chromium.org> Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
2020-06-07 20:29:18 +08:00
ret = LZ4_decompress_generic(in, out, block_size,
end - out, endOnInputSize,
decode_full_block, noDict, out, NULL, 0);
if (ret < 0) {
ret = -EPROTO; /* decompression error */
break;
}
out += ret;
}
lz4: fix decompressor on big-endian powerpc Booting an lz4-compressed kernel image fails on our powerpc board with -EPROTONOSUPPORT. Adding a bit of debug prints, we get magic: 0x184d2204 flags: 0x64 reserved0: 1 has_content_checksum: 1 has_content_size: 0 has_block_checksum: 0 independent_blocks: 1 version: 0 block_descriptor: 70 reserved1: 7 max_block_size: 0 reserved2: 0 So the magic is ok, but the version check fails, also some reserved bits are apparently set. But that's because the code interprets the "flags" and "block_descriptor" bytes wrongly: Using bit-fields to access individual bits of an "on the wire" format is not portable, not even when restricted to the C flavour implemented by gcc. Quoting the gcc manual: * 'The order of allocation of bit-fields within a unit (C90 6.5.2.1, C99 and C11 6.7.2.1).' Determined by ABI. and indeed, the PPC Processor ABI supplement says * Bit-fields are allocated from right to left (least to most significant) on Little-Endian implementations and from left to right (most to least significant) on Big-Endian implementations. The upstream code (github.com/lz4/lz4) uses explicit shifts and masks for encoding/decoding: /* FLG Byte */ *dstPtr++ = (BYTE)(((1 & _2BITS) << 6) /* Version('01') */ + ((cctxPtr->prefs.frameInfo.blockMode & _1BIT ) << 5) + ((cctxPtr->prefs.frameInfo.blockChecksumFlag & _1BIT ) << 4) + ((unsigned)(cctxPtr->prefs.frameInfo.contentSize > 0) << 3) + ((cctxPtr->prefs.frameInfo.contentChecksumFlag & _1BIT ) << 2) + (cctxPtr->prefs.frameInfo.dictID > 0) ); /* Flags */ { U32 const FLG = srcPtr[4]; U32 const version = (FLG>>6) & _2BITS; blockChecksumFlag = (FLG>>4) & _1BIT; blockMode = (FLG>>5) & _1BIT; contentSizeFlag = (FLG>>3) & _1BIT; contentChecksumFlag = (FLG>>2) & _1BIT; dictIDFlag = FLG & _1BIT; /* validate */ if (((FLG>>1)&_1BIT) != 0) return err0r(LZ4F_ERROR_reservedFlag_set); /* Reserved bit */ if (version != 1) return err0r(LZ4F_ERROR_headerVersion_wrong); /* Version Number, only supported value */ } Do the same here, and while at it, be more careful to use unaligned accessors to what is most likely unaligned. Also update the comment to make it clear that it only refers to the lz4.c file, not the following code of lz4_wrapper.c. This has been tested partly, of course, by seeing that my lz4-compressed kernel now boots, partly by running the (de)compression test-suite in the (x86_64) sandbox - i.e., it should still work just fine on little-endian hosts. Reviewed-by: Julius Werner <jwerner@chromium.org> Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
2020-06-07 20:29:18 +08:00
in += block_size;
if (has_block_checksum)
in += sizeof(u32);
}
*dstn = out - dst;
return ret;
}