From patchwork Thu Mar 26 05:35:53 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Qu Wenruo X-Patchwork-Id: 244294 List-Id: U-Boot discussion From: wqu at suse.com (Qu Wenruo) Date: Thu, 26 Mar 2020 13:35:53 +0800 Subject: [PATCH U-BOOT v2 0/3] fs: btrfs: Fix false LZO decompression error due to missing page boundary check Message-ID: <20200326053556.20492-1-wqu@suse.com> There is a bug that uboot can't load LZO compressed data extent while kernel can handle it without any problem. It turns out to be a page boundary case. The 3nd patch is the proper fix, cross-ported from btrfs-progs. The first patch is just to make my eyes less hurt. The second patch is to make sure the driver will reject sector size not matching PAGE_SIZE. This keeps the behavior the same as kernel, even in theory we could do better in U-boot. This is just a temporary fix, before better btrfs driver implemented. I guess it's time to backport proper code from btrfs-progs, other than using tons of immediate codes. Changelog: v2: - Fix code style problems - Add a new patch to reject non-page-sized sector size Since kernel does the same thing, and non-4K page size u-boot boards are really rare, it shouldn't be a big problem. Qu Wenruo (3): fs: btrfs: Use LZO_LEN to replace immediate number fs: btrfs: Reject fs with sector size other than PAGE_SIZE fs: btrfs: Fix LZO false decompression error caused by pending zero fs/btrfs/compression.c | 42 +++++++++++++++++++++++++++++++----------- fs/btrfs/super.c | 8 ++++++++ 2 files changed, 39 insertions(+), 11 deletions(-)