Skip to content

Don't drop in-flight decode jobs when seeking in cram_next_slice() #1973

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 2 commits into from
Dec 1, 2025
Merged

Don't drop in-flight decode jobs when seeking in cram_next_slice() #1973

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
4f4e16e
Don't drop in-flight decode jobs when seeking in cram_next_slice()
daviesrob Nov 25, 2025
4b53c3a
Add CRAM container skipping test
daviesrob Nov 25, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion Makefile
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -526,7 +526,7 @@ realn.o realn.pico: realn.c config.h $(htslib_hts_h) $(htslib_sam_h)
textutils.o textutils.pico: textutils.c config.h $(htslib_hfile_h) $(htslib_kstring_h) $(htslib_sam_h) $(hts_internal_h)

cram/cram_codecs.o cram/cram_codecs.pico: cram/cram_codecs.c config.h $(fuzz_settings_h) $(htslib_hts_endian_h) $(htscodecs_varint_h) $(htscodecs_pack_h) $(htscodecs_rle_h) $(cram_h)
cram/cram_decode.o cram/cram_decode.pico: cram/cram_decode.c config.h $(cram_h) $(cram_os_h) $(htslib_hts_h)
cram/cram_decode.o cram/cram_decode.pico: cram/cram_decode.c config.h $(cram_h) $(cram_os_h) $(htslib_hts_h) $(htslib_hfile_h)
cram/cram_encode.o cram/cram_encode.pico: cram/cram_encode.c config.h $(cram_h) $(cram_os_h) $(sam_internal_h) $(htslib_hts_h) $(htslib_hts_endian_h) $(textutils_internal_h)
cram/cram_external.o cram/cram_external.pico: cram/cram_external.c config.h $(htscodecs_rANS_static4x16_h) $(htslib_hfile_h) $(cram_h)
cram/cram_index.o cram/cram_index.pico: cram/cram_index.c config.h $(htslib_bgzf_h) $(htslib_hfile_h) $(hts_internal_h) $(cram_h) $(cram_os_h)
Expand Down
11 changes: 8 additions & 3 deletions cram/cram_decode.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -51,6 +51,7 @@ OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#include "cram.h"
#include "os.h"
#include "../htslib/hts.h"
#include "../htslib/hfile.h"

//Whether CIGAR has just M or uses = and X to indicate match and mismatch
//#define USE_X
Expand Down Expand Up @@ -3287,14 +3288,18 @@ cram_slice *cram_next_slice(cram_fd *fd, cram_container **cp) {
break;
}

// before start of range; skip to next container
// Before start of range; skip to next container.
// Can't use cram_seek() here as it drops in-progress
// multi-threaded decode jobs, so call hseek() directly.
if (fd->range.refid != -1 &&
c_next->ref_seq_start + c_next->ref_seq_span-1 <
fd->range.start) {
c_next->curr_slice_mt = c_next->max_slice;
cram_seek(fd, c_next->length, SEEK_CUR);
off_t skip_length = c_next->length;
cram_free_container(c_next);
c_next = NULL;
fd->ooc = 0;
if (hseek(fd->fp, skip_length, SEEK_CUR) < 0)
return NULL;
continue;
}
}
Expand Down
17 changes: 17 additions & 0 deletions test/index3.sam
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,17 @@
@HD VN:1.6 SO:coordinate
@SQ SN:CHROMOSOME_I LN:1009800 M5:8ede36131e0dbf3417807e48f77f3ebd
@CO CRAM container skipping test - very long reads followed by short ones.
@CO Use option seqs_per_slice=2 when writing CRAM so that records are
@CO stored in multiple containers.
r0-1 0 CHROMOSOME_I 1 50 10000M * 0 0 * *
r0-2 0 CHROMOSOME_I 1 50 10000M * 0 0 * *
r1-1 0 CHROMOSOME_I 1000 50 100M * 0 0 * *
r1-2 0 CHROMOSOME_I 1000 50 100M * 0 0 * *
r2-1 0 CHROMOSOME_I 2000 50 100M * 0 0 * *
r2-2 0 CHROMOSOME_I 2000 50 100M * 0 0 * *
r3-1 0 CHROMOSOME_I 3000 50 100M * 0 0 * *
r3-2 0 CHROMOSOME_I 3000 50 100M * 0 0 * *
r4-1 0 CHROMOSOME_I 4000 50 100M * 0 0 * *
r4-2 0 CHROMOSOME_I 4000 50 100M * 0 0 * *
r5-1 0 CHROMOSOME_I 5000 50 100M * 0 0 * *
r5-2 0 CHROMOSOME_I 5000 50 100M * 0 0 * *
9 changes: 9 additions & 0 deletions test/index3_exp.sam
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,9 @@
@HD VN:1.6 SO:coordinate
@SQ SN:CHROMOSOME_I LN:1009800 M5:8ede36131e0dbf3417807e48f77f3ebd
@CO CRAM container skipping test - very long reads followed by short ones.
@CO Use option seqs_per_slice=2 when writing CRAM so that records are
@CO stored in multiple containers.
r0-1 0 CHROMOSOME_I 1 50 10000M * 0 0 * *
r0-2 0 CHROMOSOME_I 1 50 10000M * 0 0 * *
r5-1 0 CHROMOSOME_I 5000 50 100M * 0 0 * *
r5-2 0 CHROMOSOME_I 5000 50 100M * 0 0 * *
12 changes: 12 additions & 0 deletions test/test.pl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1089,6 +1089,18 @@ sub test_index
unlink("$$opts{tmp}/index.cram.crai");
test_compare($opts,"$$opts{path}/test_index $$opts{tmp}/index.cram", "$$opts{tmp}/index.cram.crai", "$$opts{path}/index.cram.crai", gz=>1);

# CRAM container skipping test
# Prepare a file with records split into multiple containers, the first two
# records spanning far enough to overlap the last two.
cmd("$$opts{path}/test_view $nthreads -C -p $$opts{tmp}/index3.cram -x $$opts{tmp}/index3.cram.crai -o seqs_per_slice=2 $$opts{path}/index3.sam");
# An index lookup for the last two records should return the first two
# as well.
test_compare($opts, "$$opts{path}/test_view $nthreads -p $$opts{tmp}/index3_rgn.sam $$opts{tmp}/index3.cram CHROMOSOME_I:5000-5100",
"$$opts{path}/index3_exp.sam", "$$opts{tmp}/index3_rgn.sam");
# Also check the multi-region iterator
test_compare($opts, "$$opts{path}/test_view $nthreads -M -p $$opts{tmp}/index3_rgnm.sam $$opts{tmp}/index3.cram CHROMOSOME_I:5000-5100",
"$$opts{path}/index3_exp.sam", "$$opts{tmp}/index3_rgnm.sam");

# BCF
test_compare($opts,"$$opts{path}/test_view $nthreads -l 0 -b -m 14 -x $$opts{tmp}/index.bcf.csi $$opts{path}/index.vcf > $$opts{tmp}/index.bcf", "$$opts{tmp}/index.bcf.csi", "$$opts{path}/index.bcf.csi", gz=>1);
unlink("$$opts{tmp}/index.bcf.csi");
Expand Down