diff --git a/SOURCES/bz1819518-fix-handling-hpack-zero-bytes-overwrite.patch b/SOURCES/bz1819518-fix-handling-hpack-zero-bytes-overwrite.patch new file mode 100644 index 0000000..24ed204 --- /dev/null +++ b/SOURCES/bz1819518-fix-handling-hpack-zero-bytes-overwrite.patch @@ -0,0 +1,51 @@ +From 4e372dc350be5c72b88546bf03392a5793cea179 Mon Sep 17 00:00:00 2001 +From: Willy Tarreau +Date: Sun, 29 Mar 2020 08:53:31 +0200 +Subject: BUG/CRITICAL: hpack: never index a header into the headroom after + wrapping + +The HPACK header table is implemented as a wrapping list inside a contigous +area. Headers names and values are stored from right to left while indexes +are stored from left to right. When there's no more room to store a new one, +we wrap to the right again, or possibly defragment it if needed. The condition +do use the right part (called tailroom) or the left part (called headroom) +depends on the location of the last inserted header. After wrapping happens, +the code forces to stick to tailroom by pretending there's no more headroom, +so that the size fit test always fails. The problem is that nothing prevents +from storing a header with an empty name and empty value, resulting in a +total size of zero bytes, which satisfies the condition to use the headroom. +Doing this in a wrapped buffer results in changing the "front" header index +and causing miscalculations on the available size and the addresses of the +next headers. This may even allow to overwrite some parts of the index, +opening the possibility to perform arbitrary writes into a 32-bit relative +address space. + +This patch fixes the issue by making sure the headroom is considered only +when the buffer does not wrap, instead of relying on the zero size. This +must be backported to all versions supporting H2, which is as far as 1.8. + +Many thanks to Felix Wilhelm of Google Project Zero for responsibly +reporting this problem with a reproducer and a detailed analysis. +--- + src/hpack-tbl.c | 4 ++-- + 1 file changed, 2 insertions(+), 2 deletions(-) + +diff --git a/src/hpack-tbl.c b/src/hpack-tbl.c +index 70d7f35834..727ff7a17b 100644 +--- a/src/hpack-tbl.c ++++ b/src/hpack-tbl.c +@@ -346,9 +346,9 @@ int hpack_dht_insert(struct hpack_dht *dht, struct ist name, struct ist value) + * room left in the tail to suit the protocol, but tests show that in + * practice it almost never happens in other situations so the extra + * test is useless and we simply fill the headroom as long as it's +- * available. ++ * available and we don't wrap. + */ +- if (headroom >= name.len + value.len) { ++ if (prev == dht->front && headroom >= name.len + value.len) { + /* install upfront and update ->front */ + dht->dte[head].addr = dht->dte[dht->front].addr - (name.len + value.len); + dht->front = head; +-- +2.20.1 + diff --git a/SPECS/haproxy.spec b/SPECS/haproxy.spec index 490d006..c3efda6 100644 --- a/SPECS/haproxy.spec +++ b/SPECS/haproxy.spec @@ -8,7 +8,7 @@ Name: haproxy Version: 1.8.15 -Release: 5%{?dist} +Release: 6%{?dist}.1 Summary: HAProxy reverse proxy for high availability environments Group: System Environment/Daemons @@ -23,6 +23,7 @@ Source4: %{name}.sysconfig Source5: halog.1 Patch0: bz1664533-fix-handling-priority-flag-HTTP2-decoder.patch +Patch1: bz1819518-fix-handling-hpack-zero-bytes-overwrite.patch BuildRequires: lua-devel BuildRequires: pcre-devel @@ -53,6 +54,7 @@ availability environments. Indeed, it can: %prep %setup -q %patch0 -p1 +%patch1 -p1 %build regparm_opts= @@ -138,6 +140,12 @@ exit 0 %{_mandir}/man1/* %changelog +* Wed Apr 01 2020 Ryan O'Hara - 1.8.15-6.1 +- - Fix hapack zero byte input causing overwrite (CVE-2020-11100, #1819518) + +* Fri Jul 19 2019 Ryan O'Hara - 1.8.15-6 +- Add gating tests (#1682106) + * Wed Jan 09 2019 Ryan O'Hara - 1.8.15-5 - Resolve CVE-2018-20615 (#1664533)