Created attachment 1708413 [details]
build.log

file build.log too big, will only attach last 32768 bytes

Created attachment 1708414 [details]
root.log

file root.log too big, will only attach last 32768 bytes

Created attachment 1708415 [details]
state.log

Looks like this is caused by the recent LTO enablement, though the actual errors look like a false-positive:

 alg1485.c: In function 'AppendAVA.constprop':
 alg1485.c:1049:21: error: writing 1 byte into a region of size 0 [-Werror=stringop-overflow=]
 alg1485.c:928:10: note: at offset -1 to object 'tmpBuf' with size 2048 declared here
 alg1485.c:1050:21: error: writing 1 byte into a region of size 0 [-Werror=stringop-overflow=]
 alg1485.c:928:10: note: at offset -2 to object 'tmpBuf' with size 2048 declared here
 alg1485.c:1051:21: error: writing 1 byte into a region of size 0 [-Werror=stringop-overflow=]
 alg1485.c:928:10: note: at offset -3 to object 'tmpBuf' with size 2048 declared here

where tmpBuf is accessed with something like:

 char tmpBuf[TMPBUF_LEN];
 char *end = encodedAVA /* points to tmpBuf */ + nameLen + valueLen;
 end[-1] = '.';

Perhaps we can work it around with explicitly specifying the positive index.

Let me take this in the immediate term.  I'll disable LTO in the immediate term and pull together a testcase for the false positive as that's something we'll want to dig into upstream in GCC.

This bug appears to have been reported against 'rawhide' during the Fedora 33 development cycle.
Changing version to 33.

I've disabled LTO for nss.  But further work is needed as I'm still seeing failures:

dbmshim.c:8:10: fatal error: mcom_db.h: No such file or directory

The package owner needs to take it from here.

Yes, should be fixed now.