[llvm-commits] CVS: llvm-www/devmtg/2009-10/index.html

[Tanya Lattner]

Changes in directory llvm-www/devmtg/2009-10:

index.html updated: 1.5 -> 1.6
---
Log message:

Shorten descriptions.


---
Diffs of the changes:  (+6 -5)

 index.html |   11 ++++++-----
 1 files changed, 6 insertions(+), 5 deletions(-)


Index: llvm-www/devmtg/2009-10/index.html
diff -u llvm-www/devmtg/2009-10/index.html:1.5 llvm-www/devmtg/2009-10/index.html:1.6
--- llvm-www/devmtg/2009-10/index.html:1.5	Mon Oct 12 17:03:46 2009
+++ llvm-www/devmtg/2009-10/index.html	Mon Oct 12 17:08:39 2009
@@ -88,7 +88,7 @@
     </td>
     <td>Nicolas Geoffray<br><i>Universite Pierre et Marie Curie</i></td>
     <td><b>Precise and efficient garbage collection in VMKit with MMTk</b> - 
-    This talk will describe how we have added precise garbage collection in VMKit, both in the JVM and .Net runtime. The work continues in the spirit of VMKit, in that the garbage collector is an external library provided by others. We have used MMTk, a garbage collector written in Java to provide efficient and precise garbage collection. The talk will give a step by step explanation on how the new system was implemented (using LLVM as an ahead of time compiler for Java, using the LLVM GC framework, creating stack maps at runtime, etc), and provide a tutorial on how this can be adopted by other kinds of VMs implemented with LLVM. Additionally, I will show the new performance results with the new garbage collector. The new system continues to support big applications (eclipse, tomcat, DaCapo benchmarks), and provides way better performance than before.
+    This talk will describe how we have added precise garbage collection in VMKit, both in the JVM and .Net runtime. The work continues in the spirit of VMKit, in that the garbage collector is an external library provided by others. We have used MMTk, a garbage collector written in Java to provide efficient and precise garbage collection. The talk will give a step by step explanation on how the new system was implemented and provide a tutorial on how this can be adopted by other kinds of VMs implemented with LLVM. Additionally, I will show the new performance results with the new garbage collector. 
     </td>
   </tr>
   
@@ -187,8 +187,9 @@
       [<a href="Zaks_CoVaC.pdf">Slides</a>]
     </td>
     <td>Anna Zaks<br><i>New York University</i></td>
-    <td><b>CoVaC: Compiler Validation by Program Analysis of the Cross-Product</b> - I 
-    n this talk, I am going to present CoVaC, which is a framework for formal verification of the compiler optimization phase, and give a brief overview of the prototype tool, which has been developed on top of LLVM 2.0. The CoVaC framework checks that optimization passes preserve semantics of the program being compiled by proving that the unoptimized program is equivalent to the optimized one. To leverage existing program analysis techniques, we reduce the equivalence checking problem to analysis of one system - a cross-product of the two input programs. Unlike existing frameworks, CoVaC accommodates absence of compiler annotations and handles most of the classical intraprocedural optimizations such as constant folding, reassociation, common subexpression elimination, code motion, dead code elimination, branch optimizations, and others. The talk is based on the following paper: Compiler Validation by Program Analysis of the Cross-Product, Anna Zaks and Amir Pnueli, FM'08
+    <td><b>CoVaC: Compiler Validation by Program Analysis of the Cross-Product</b> -  
+    CoVaC is a framework for formal verification of the compiler optimization phase. This talk gives a brief overview of the prototype tool, which has been developed on top of LLVM 2.0. The CoVaC framework checks that optimization passes preserve semantics of the program being compiled by proving that the unoptimized program is equivalent to the optimized one. To leverage existing program analysis techniques, we reduce the equivalence checking problem to analysis of one system - a cross-product of the two input programs. 
+The talk is based on the following paper: Compiler Validation by Program Analysis of the Cross-Product, Anna Zaks and Amir Pnueli, FM'08
     </td>
   </tr>
 
@@ -208,7 +209,7 @@
     </td>
     <td>Santosh Nagarakatte<br><i>University of Pennsylvania</i></td>
     <td><b>SoftBound: Highly Compatible and Complete Spatial Memory Safety for C</b> - 
-    The serious bugs and security vulnerabilities facilitated by C/C++’s lack of bounds checking are well known, yet C and C++ remain in widespread use. Unfortunately, C’s arbitrary pointer arithmetic, conflation of pointers and arrays, and programmer-visible memory layout make retrofitting C/C++ with spatial safety guarantees extremely challenging. Existing approaches suffer from incompleteness, have high runtime overhead, or require non-trivial changes to the C source code. Thus far, these deficiencies have prevented widespread adoption of such techniques. In this talk, I will present SoftBound, a compile-time transformation for enforcing spatial safety of C. SoftBound records base and bound information for every pointer as disjoint metadata. This decoupling enables SoftBound to provide spatial safety without requiring changes to C source code. SoftBound is a software-only approach and performs metadata manipulation only when loading or storing pointer values. I will !
 also provide a brief description of the formal proof and llvm implementation. SoftBound’s full checking mode provides complete spatial violation detection with 67% runtime overhead on average. To further reduce overheads, SoftBound has a store-only checking mode that successfully detects all the security vulnerabilities in a test suite at the cost of only 22% runtime overhead on average.
+SoftBound is a compile-time transformation for enforcing spatial safety of C. SoftBound records base and bound information for every pointer as disjoint metadata. This decoupling enables SoftBound to provide spatial safety without requiring changes to C source code. SoftBound is a software-only approach and performs metadata manipulation only when loading or storing pointer values. This talk provides a brief description of the formal proof and llvm implementation. 
     </td>
   </tr>
 
@@ -249,7 +250,7 @@
   src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
   <a href="http://validator.w3.org/check/referer"><img
   src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!"></a>
-<br>Last modified: $Date: 2009/10/12 22:03:46 $
+<br>Last modified: $Date: 2009/10/12 22:08:39 $
 </address>
 
 <!--#include virtual="../../footer.incl" -->