Reduce ram used for very large inlined index files
During a clone, an inline index is not converted to a split index
file until the very end. When the conversion happens, the index
can be very large, and the inline index loading functions always load
the entire index file into ram.
This changes the revlog code to read the index in smaller chunks.
#!/bin/sh
mkdir t
cd t
hg init
echo "added file1" > file1
echo "another line of text" >> file1
echo "added file2" > file2
hg add file1 file2
hg commit -m "added file1 and file2" -d "1000000 0" -u user
echo "changed file1" >> file1
hg commit -m "changed file1" -d "1000000 0" -u user
hg -q log
hg id
hg update -C 0
hg id
echo "changed file1" >> file1
hg id
hg revert --no-backup
hg diff
hg status
hg id
hg update
hg diff
hg status
hg id
hg update -C 0
echo "changed file1 different" >> file1
HGMERGE=merge hg update
hg diff | sed -e "s/\(+++ [a-zA-Z0-9_/.-]*\).*/\1/" -e "s/\(<<<<<<<\) .*/\1/" \
-e "s/\(--- [a-zA-Z0-9_/.-]*\).*/\1/" -e "s/\(>>>>>>>\) .*/\1/"
hg status
hg id
hg revert --no-backup
hg diff
hg status
hg id
hg revert -r tip --no-backup
hg diff
hg status
hg id
hg update -C
hg diff
hg status
hg id
cd ..; /bin/rm -rf t