From 867ea1ee3fd02e15487c308dc9e5ef2bb141f7c2 Mon Sep 17 00:00:00 2001
From: Denys Vlasenko <vda.linux@googlemail.com>
Date: Mon, 22 Feb 2010 13:49:22 +0100
Subject: Kerneloops: make hashing more likely to produce same hash on
 different oopses

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
---
 lib/Plugins/Kerneloops.cpp | 83 ++++++++++++++++++++++++++++++++++++++++------
 1 file changed, 73 insertions(+), 10 deletions(-)

(limited to 'lib')

diff --git a/lib/Plugins/Kerneloops.cpp b/lib/Plugins/Kerneloops.cpp
index 12ce3783..fb7eaca7 100644
--- a/lib/Plugins/Kerneloops.cpp
+++ b/lib/Plugins/Kerneloops.cpp
@@ -23,6 +23,78 @@
 #include "ABRTException.h"
 #include "CommLayerInner.h"
 
+static unsigned hash_oops_str(const char *oops_ptr)
+{
+	unsigned char old_c;
+	unsigned char c = 0;
+	unsigned hash = 0;
+	while (1)
+	{
+		old_c = c;
+		c = *oops_ptr++;
+		if (!c)
+			break;
+		if (c == '\n') {
+			// Exclude some lines which have process name - in some oops classes
+			// process name is irrelevant and changes with every oops.
+			// Lines we filter out:
+			// Pid: 8003, comm: Xorg Not tainted (2.6.27.9-159.fc10.i686 #1)
+			// Process Xorg (pid: 8003, ti=f0a0c000 task=f2380000 task.ti=f0a0c000)
+			if (strncmp(oops_ptr, "Pid: ", 5) == 0
+			 || strncmp(oops_ptr, "Process ", 8) == 0
+			) {
+				while (*oops_ptr && *oops_ptr != '\n')
+					oops_ptr++;
+				continue;
+			}
+		}
+		if (!isalnum(old_c)) {
+			if (c >= '0' && c <= '9') {
+				// Convert all (possibly hex) numbers to just one '0'
+				if (c == '0' && *oops_ptr == 'x') // "0xSOMETHING"
+					oops_ptr++;
+				while (isxdigit(*oops_ptr))
+					oops_ptr++;
+				c = '0';
+			} else
+			if ((c|0x20) >= 'a' && (c|0x20) <= 'f') {
+				// This *may be* a hex number without 0x prefix: "f0a0c000"
+				// Check that it indeed is, and replace with '0'
+				const char *oops_ptr2 = oops_ptr;
+				while (isxdigit(*oops_ptr2))
+					oops_ptr2++;
+				// Does it end in a letter which is not a hex digit?
+				// (Example: "abcw" is not a hex number, "abc " is)
+				if (!isalpha(*oops_ptr2)) {
+					// It's "abc " case. Skip the "abc" string
+					oops_ptr = oops_ptr2;
+					c = '0';
+				}
+				// else: hash the string as-is
+			}
+		}
+		// TODO: Drop call trace tail - in interrupt-driven oopses,
+		// everything before interrupt is irrelevant.
+		// Example of call trace part of oops:
+		// Call Trace:
+		// [<f88e11c7>] ? radeon_cp_resume+0x7d/0xbc [radeon]
+		// [<f88745f8>] ? drm_ioctl+0x1b0/0x225 [drm]
+		// [<f88e114a>] ? radeon_cp_resume+0x0/0xbc [radeon]
+		// [<c049b1c0>] ? vfs_ioctl+0x50/0x69
+		// [<c049b414>] ? do_vfs_ioctl+0x23b/0x247
+		// [<c0460a56>] ? audit_syscall_entry+0xf9/0x123
+		// [<c049b460>] ? sys_ioctl+0x40/0x5c
+		// [<c0403c76>] ? syscall_call+0x7/0xb
+
+		/* An algorithm proposed by Donald E. Knuth in The Art Of Computer
+		 * Programming Volume 3, under the topic of sorting and search
+		 * chapter 6.4.
+		 */
+		hash = ((hash << 5) ^ (hash >> 27)) ^ c;
+	}
+	return hash;
+}
+
 std::string CAnalyzerKerneloops::GetLocalUUID(const char *pDebugDumpDir)
 {
 	log(_("Getting local universal unique identification"));
@@ -34,16 +106,7 @@ std::string CAnalyzerKerneloops::GetLocalUUID(const char *pDebugDumpDir)
 		dd.LoadText(FILENAME_KERNELOOPS, oops);
 	}
 
-	/* An algorithm proposed by Donald E. Knuth in The Art Of Computer
-	 * Programming Volume 3, under the topic of sorting and search
-	 * chapter 6.4.
-	 */
-	unsigned len = oops.length();
-	unsigned hash = len;
-	for (unsigned i = 0; i < len; i++)
-	{
-		hash = ((hash << 5) ^ (hash >> 27)) ^ oops[i];
-	}
+	unsigned hash = hash_oops_str(oops.c_str());
 	hash &= 0x7FFFFFFF;
 
 	return to_string(hash);
-- 
cgit