VSEARCH 1.0.13: Fix for memory bug with multiple alignments and clust…

…ering
torognes · Feb 17, 2015 · 9c0f3a1 · 9c0f3a1
1 parent 0bb8fb1
commit 9c0f3a1
Show file tree

Hide file tree

Showing 8 changed files with 16 additions and 17 deletions.
diff --git a/README.md b/README.md
@@ -32,22 +32,22 @@ If you can't find an answer in the VSEARCH documentation, please visit the [VSEA
 
 In the example below, VSEARCH will identify sequences in the file database.fsa that are at least 90% identical on the plus strand to the query sequences in the file queries.fsa and write the results to the file alnout.txt.
 
-`./vsearch-1.0.11-linux-x86_64 --usearch_global queries.fsa --db database.fsa --id 0.9 --alnout alnout.txt`
+`./vsearch-1.0.13-linux-x86_64 --usearch_global queries.fsa --db database.fsa --id 0.9 --alnout alnout.txt`
 
 ## Download and install
 
 The latest releases of VSEARCH are available [here](https://github.com/torognes/vsearch/releases).
 
-Binary executables of VSEARCH are available in the `bin` folder for [GNU/Linux on x86-64 systems](https://github.com/torognes/vsearch/blob/master/bin/vsearch-1.0.11-linux-x86_64) and [Apple Mac OS X on x86-64 systems](https://github.com/torognes/vsearch/blob/master/bin/vsearch-1.0.11-osx-x86_64). These executables include support for  input files compressed by zlib and bzip2 (with files usually ending in .gz or .bz2).
+Binary executables of VSEARCH are available in the `bin` folder for [GNU/Linux on x86-64 systems](https://github.com/torognes/vsearch/blob/master/bin/vsearch-1.0.13-linux-x86_64) and [Apple Mac OS X on x86-64 systems](https://github.com/torognes/vsearch/blob/master/bin/vsearch-1.0.13-osx-x86_64). These executables include support for  input files compressed by zlib and bzip2 (with files usually ending in .gz or .bz2).
 
 Download the appropriate executable and make a symbolic link in a folder included in your `$PATH` from `vsearch` to the appropriate binary. You may use the following commands (assuming `~/bin` is in your `$PATH`):
 
 ```sh
 cd ~
 mkdir -p bin
 cd bin
-wget https://github.com/torognes/vsearch/releases/download/v1.0.11/vsearch-1.0.11-linux-x86_64
-ln -s vsearch-1.0.11-linux-x86_64 vsearch
+wget https://github.com/torognes/vsearch/releases/download/v1.0.13/vsearch-1.0.13-linux-x86_64
+ln -s vsearch-1.0.13-linux-x86_64 vsearch
 ```
 
 Substitute `linux` with `osx` in those lines if you're on a Mac.

diff --git a/doc/vsearch.1 b/doc/vsearch.1
@@ -1,5 +1,5 @@
 .\" ============================================================================
-.TH vsearch 1 "February 6, 2015" "version 1.0.12" "USER COMMANDS"
+.TH vsearch 1 "February 17, 2015" "version 1.0.13" "USER COMMANDS"
 .\" ============================================================================
 .SH NAME
 vsearch \(em chimera detection, clustering, dereplication, masking, pairwise alignment, searching, shuffling and sorting of amplicons from metagenomic projects.
@@ -1523,6 +1523,11 @@ clustering, pairwise alignment and searching).
 .TP
 .BR v1.0.12\~ "released February 6th, 2015"
 Temporarily fixes a problem with long headers in FASTA files.
+.TP
+.BR v1.0.13\~ "released February 6th, 2015"
+Fix a memory allocation problem when computing multiple sequence alignments
+with the --msaout and --consout options, as well as a memory leak.
+Also increased line buffer for reading FASTA files to 4MB.
 .LP
 .\" ============================================================================
 .\" TODO:

diff --git a/doc/vsearch_manual.pdf b/doc/vsearch_manual.pdf
diff --git a/src/cluster.cc b/src/cluster.cc
@@ -1110,8 +1110,6 @@ void cluster(char * dbname,
       int msa_target_count = 0;
       struct msa_target_s * msa_target_list =
         (struct msa_target_s *) xmalloc(sizeof(struct msa_target_s) * size_max);
-      //memset(msa_target_list, 0, sizeof(struct msa_target_s)*size_max);
-
       progress_init("Multiple alignments", seqcount);
 
       FILE * fp_msaout = 0;
@@ -1168,11 +1166,11 @@ void cluster(char * dbname,
     }
 
   /* free cigar strings for all aligned sequences */
-  /*
+
   for(int i=0; i<seqcount; i++)
     if (clusterinfo[i].cigar)
       free(clusterinfo[i].cigar);
-  */
+
   free(clusterinfo);
 
   if (opt_matched)

diff --git a/src/db.cc b/src/db.cc
@@ -34,7 +34,7 @@ seqinfo_t * seqindex = 0;
 static char * datap = 0;
 
 #define MEMCHUNK 10485760
-#define LINEALLOC 1048576
+#define LINEALLOC 4194304
 
 regex_t db_regexp;
 

diff --git a/src/msa.cc b/src/msa.cc
@@ -1,5 +1,5 @@
 /*
-    Copyright (C) 2014 Torbjorn Rognes
+    Copyright (C) 2014-2015 Torbjorn Rognes
 
     This program is free software: you can redistribute it and/or modify
     it under the terms of the GNU Affero General Public License as
@@ -233,10 +233,6 @@ void msa(FILE * fp_msaout, FILE * fp_consout,
                   db_getheader(target_seqno));
           fprint_fasta_seq_only(fp_msaout, aln, alnlen, opt_fasta_width);
         }
-
-      /* free cigar */
-      if (target_list[j].cigar)
-        free(target_list[j].cigar);
     }  
 
 

diff --git a/src/query.cc b/src/query.cc
@@ -26,7 +26,7 @@
    be overwritten on the next call of query_getnext. */
 
 #define MEMCHUNK 4096
-#define LINEALLOC 1048576
+#define LINEALLOC 4194304
 
 extern unsigned int chrstatus[256];
 

diff --git a/src/vsearch.h b/src/vsearch.h
@@ -84,7 +84,7 @@
 #include "allpairs.h"
 
 #define PROG_NAME "vsearch"
-#define PROG_VERSION "v1.0.12"
+#define PROG_VERSION "v1.0.13"
 
 #ifdef __APPLE__
 #define PROG_ARCH "osx_x86_64"