X-Git-Url: https://git.lttng.org/?a=blobdiff_plain;f=ltt%2Fbranches%2Fpoly%2Fdoc%2Fdeveloper%2Flttng-lttv-compatibility.html;h=f0c3a90c7004afdf7024eda79e95c876aa03414f;hb=f4f60f519f12a1de11514ddf6f84f0c7d69ce2d2;hp=dc047e82f5877a0c2d2daabed13b39d1b9106bda;hpb=775c18a26cffc0ef09529bc2bc471c72930a634f;p=lttv.git

diff --git a/ltt/branches/poly/doc/developer/lttng-lttv-compatibility.html b/ltt/branches/poly/doc/developer/lttng-lttv-compatibility.html
index dc047e82..f0c3a90c 100644
--- a/ltt/branches/poly/doc/developer/lttng-lttv-compatibility.html
+++ b/ltt/branches/poly/doc/developer/lttng-lttv-compatibility.html
@@ -28,7 +28,7 @@ cellspacing="2" rules="all" frame="border">
 </th>
 <th style="vertical-align: top;">Comments<br>
 </th>
-<th style="vertical-align: top;">Regression ok<br>
+<th style="vertical-align: top;">Kernel Regression ok<br>
 </th>
 <th style="vertical-align: top;">Regression fails/not tested<br>
 </th>
@@ -54,7 +54,7 @@ cellspacing="2" rules="all" frame="border">
 </th>
 <th style="vertical-align: top;">Comments<br>
 </th>
-<th style="vertical-align: top;">Regression ok<br>
+<th style="vertical-align: top;">Kernel Regression ok<br>
 </th>
 <th style="vertical-align: top;">Regression fails/not tested<br>
 </th>
@@ -2110,7 +2110,6 @@ Use cpuid instruction in pIII XMC.<br>
 0.8.77<br>
 0.8.78<br>
 0.8.79<br>
-0.8.81<br>
 </td>
 <td style="vertical-align: top;">
 0.6.62<br>
@@ -2187,9 +2186,11 @@ xtensa.<br>
 <tr>
 <td style="vertical-align: top;">
 0.8.80<br>
+0.8.81<br>
 </td>
 <td style="vertical-align: top;">
 0.6.78<br>
+0.6.80<br>
 </td>
 <td style="vertical-align: top;">
 0.36<br>
@@ -2215,6 +2216,10 @@ Add compact channel.<br>
 Fix start of trace get full timestamp. Caused problems with time gap between
 trace create/start and 32 bits (or less) TSC in events.<br>
 Ok for xen-unstable changeset starting at 14391.<br>
+Adds TSC testing in kernel.<br>
+Adds monotonic logical clock based on highest TSC count for x86 and x86_64 with
+async TSCs.<br>
+Fix missing compat_fs_exec event : rename to fs_exec.<br>
 </td>
 <td style="vertical-align: top;">
 </td>
@@ -2226,6 +2231,56 @@ xtensa.<br>
 
 </tr>
 
+<tr>
+<td style="vertical-align: top;">
+0.8.82<br>
+</td>
+<td style="vertical-align: top;">
+0.9.0<br>
+</td>
+<td style="vertical-align: top;">
+0.37<br>
+</td>
+<td style="vertical-align: top;">
+obsolete<br>
+</td>
+<td style="vertical-align: top;">
+0.29<br>
+</td>
+<td style="vertical-align: top;">
+0.34<br>
+</td>
+<td style="vertical-align: top;">
+0.8<br>
+</td>
+<td style="vertical-align: top;">
+2.6.20<br>
+ARM, MIPS32/64, powerpc32, powerpc64, ppc, i386, x86_64<br>
+</td>
+<td style="vertical-align: top;">
+Introduces the LTT serialization library. Cuts the kernel patch size from
+2M to 500k. Deprecates genevent for kernel instrumentation, but keeps backward
+compatibility with genevent generated code (still used for user space tracing).
+Slightly longer execution time at instrumentation site (270ns vs 200ns for 4
+bytes write on P4 3GHz), but the global advantage of using less memory, and
+therefore less cache, should overweight this cost in cycles. Note : In this
+version, the ltt-probe-* kernel modules and the XML definitions must be kept in
+sync by hand, otherwise, LTTV will detect a size mismatch between the kernel
+trace and the expected event size..<br>
+</td>
+<td style="vertical-align: top;">
+arm, i686, ia64, m68k, mips, mipsel, x86_64, powerpc 405, powerpc64, s390,
+sparc, sparc64.<br>
+</td>
+<td style="vertical-align: top;">
+alpha, arm26, avr32, cris, frv, h8300, m32r, m68knommu, parisc, sh, sh64, um,
+v850, xtensa.<br>
+</td>
+
+</tr>
+
+
+
 </tbody>
 </table>