makedumpfile: make reserve_diskspace do nothing for flattened format
Subsystems (2)
PATCH 1/1 makedumpfile: make reserve_diskspace do nothing for flattened format
makedumpfile: make reserve_diskspace do nothing for flattened format
reserve_diskspace() is called by write_elf_header() to make sure there is always space to write the program header, even if writing other data fails because of ENOSPC.
This is harmful when writing the flattened format to STDOUT for two reasons:
First, it actually wastes disk space, because first the block of zeroes is sent to STDOUT by reserve_diskspace() and then the actual program header is sent, meant to overwrite the zeroes when the flattened format is rearranged.
Second, the algorithm used to read flattened format directly by the crash program does not cope with the flattened file containing two chunks meant for the same offset. It uses a binary search on a sorted array of flat_data headers to find the data in the flat file. It may return the zeroed chunk written by reserve_diskspace() near the beginning of the file instead of the actual ELF header located near the end of the flattened file.
--- makedumpfile.c | 3 +++ 1 file changed, 3 insertions(+)
makedumpfile.c +3 -0
@@ -5206,6 +5206,9 @@ reserve_diskspace(int fd, off_t start_offset, off_t end_offset, char *file_name)int ret = FALSE; +if (info->flag_flatten) +return TRUE; +assert (start_offset < end_offset); buf_size = end_offset - start_offset;
Signature · 11 lines
2.43.0 -- Jiri Bohac <[email protected]> SUSE Labs, Prague, Czechia _______________________________________________ kexec mailing list [email protected] http://lists.infradead.org/mailman/listinfo/kexec
Hi Jiri,
sorry for the long delay.
Quoted 12 lines · Jiri Bohac
makedumpfile: make
reserve_diskspacedo nothing for flattened format
reserve_diskspace()is called bywrite_elf_header()to make sure there is always space to write the program header, even if writing other data fails because of ENOSPC.This is harmful when writing the flattened format to STDOUT for two reasons:
First, it actually wastes disk space, because first the block of zeroes is sent to STDOUT by
reserve_diskspace()and then the actual program header is sent, meant to overwrite the zeroes when the flattened format is rearranged.
Quoted 6 lines
Second, the algorithm used to read flattened format directly by the crash program does not cope with the flattened file containing two chunks meant for the same offset. It uses a binary search on a sorted array of
flat_dataheaders to find the data in the flat file. It may return the zeroed chunk written byreserve_diskspace()near the beginning of the file instead of the actual ELF header located near the end of the flattened file.
Thank you for the patch, I found a vmcore that reproduced the issue:
$ makedumpfile -FEd 31 vmcore > dump.FEd31 $ crash vmlinux dump.FEd31 ... realloc: No such file or directory cannot realloc resized ELF header buffer $
and the patch fixed this, so
(Masa will apply the patch, please wait for a while.)
Thanks,
Kazu
Quoted 13 lines
Fixes e39216fce9f73759509ec158e39c289e6c211125 ("Make the incomplete dumpfile generated by ENOSPC error analyzable.")Signed-off-by Jiri Bohac <[email protected]>--- makedumpfile.c | 3 +++ 1 file changed, 3 insertions(+)makedumpfile.c +0
-0
makedumpfile.c reserve_diskspace(int fd, off_t start_offset, off_t end_offset, char *file_name) · line 5214
Quoted 9 lines
makedumpfile.c +3
-0@@ -5206,6 +5206,9 @@ reserve_diskspace(int fd, off_t start_offset, off_t end_offset, char *file_name)int ret = FALSE; +if (info->flag_flatten) +return TRUE; +assert (start_offset < end_offset); buf_size = end_offset - start_offset;
_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec
Quoted 55 lines · HAGIO KAZUHITO(萩尾 一仁)
Hi Jiri,
sorry for the long delay.
Quoted 12 lines · Jiri Bohac
makedumpfile: make
reserve_diskspacedo nothing for flattened format
reserve_diskspace()is called bywrite_elf_header()to make sure there is always space to write the program header, even if writing other data fails because of ENOSPC.This is harmful when writing the flattened format to STDOUT for two reasons:
First, it actually wastes disk space, because first the block of zeroes is sent to STDOUT by
reserve_diskspace()and then the actual program header is sent, meant to overwrite the zeroes when the flattened format is rearranged.Quoted 6 lines
Second, the algorithm used to read flattened format directly by the crash program does not cope with the flattened file containing two chunks meant for the same offset. It uses a binary search on a sorted array of
flat_dataheaders to find the data in the flat file. It may return the zeroed chunk written byreserve_diskspace()near the beginning of the file instead of the actual ELF header located near the end of the flattened file.Thank you for the patch, I found a vmcore that reproduced the issue:
$ makedumpfile -FEd 31 vmcore > dump.FEd31 $ crash vmlinux dump.FEd31 ... realloc: No such file or directory cannot realloc resized ELF header buffer $and the patch fixed this, so
Acked-by Kazuhito Hagio <[email protected]>(Masa will apply the patch, please wait for a while.)
Thanks,
KazuQuoted 13 lines
Fixes e39216fce9f73759509ec158e39c289e6c211125 ("Make the incomplete dumpfile generated by ENOSPC error analyzable.")Signed-off-by Jiri Bohac <[email protected]>--- makedumpfile.c | 3 +++ 1 file changed, 3 insertions(+)makedumpfile.c +0
-0@@ -5206,6 +5206,9 @@ reserve_diskspace(int fd, off_t start_offset, off_t end_offset, char *file_name)
Quoted 8 lines
makedumpfile.c reserve_diskspace(int fd, off_t start_offset, off_t end_offset, char *file_name) · line 5214
Quoted 8 lines
makedumpfile.c +3
-0int ret = FALSE; +if (info->flag_flatten) +return TRUE; +assert (start_offset < end_offset); buf_size = end_offset - start_offset;
Thank you, Jiri, for the patch. Sorry for the late apply the patch.
https://github.com/makedumpfile/makedumpfile/commit/c76301277552b95b613ffe9d5a211d01df73bd79
Thanks,
Masa
_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec