[PATCHES] MSZIP support for cabinet files

[Aurélien Aptel]

Hi Jeremy,

This is in the push function, the checksum offset is obtained directly
from the ndr writer offset here:

Jeremy Allison via samba-technical <samba-technical at lists.samba.org>
writes:
>> +
>> +	for (i = 0; i < num_cfdata; i++, r++) {
>> +		uint32_t compressed_offset, compressed_length = 0;
>> +		uint32_t csum, csum_offset;
>> +		uint32_t csumPartial;
>> +
>> +		/*
>> +		 * checksum is a function of the size fields
>> +		 * and the potentially compressed data bytes,
>> +		 * which haven't been compressed yet so
>> +		 * remember offset, write zeroes, fill out
>> +		 * later
>> +		 */
>> +		csum_offset = ndr->offset;
>> +		NDR_CHECK(ndr_push_uint32(ndr, NDR_SCALARS, 0));
>> +


>> +		csumPartial = ndr_cab_compute_checksum(&ndr->data[csum_offset + 8],
>
>                                                                  ^^^^^^^^^^^^^^^^
>                                                                  Integer wrap checks ?
> Didn't cnum_offset come from the client ?


The expression will wrap around 4GB indeed but since many fields of CAB
files are 32 bits I'm guessing it's also the file size hard limit of a
CAB file.

Since it's the push function we should probably return an error early if
the data to serialize is too big.

>> +			/* write in the folder header the offset of its first data block */
>> +			for (i = 0; i < r->cfheader.cFolders; i++) {
>> +				SIVAL(ndr->data, OFFSET_OF_FOLDER_COFFCABSTART(i), ndr->offset);
>> +				NDR_CHECK(ndr_push_folder_cfdata(ndr, r->cfdata + processed_cfdata, r->cffolders[i].typeCompress, r->cffolders[i].cCFData));
>> +				processed_cfdata += r->cffolders[i].cCFData;
> 				^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> 				Here also ?

Same comment here I think.

-- 
Aurélien Aptel / SUSE Labs Samba Team
GPG: 1839 CB5F 9F5B FB9B AA97  8C99 03C8 A49B 521B D5D3
SUSE Linux GmbH, Maxfeldstraße 5, 90409 Nürnberg, Germany
GF: Felix Imendörffer, Jane Smithard, Graham Norton, HRB 21284 (AG Nürnberg)