Re: [9front] Cannot build system on latest rpi4 (libbio errors)

Re: [9front] Cannot build system on latest rpi4 (libbio errors)

[Romano]

[-- Attachment #1: Type: text/plain, Size: 754 bytes --]

To clarify, the failure wasn't exactly the same. Here is where it failed:

cd ../smtp && mk rfc822.tab.7
mk: 'rfc822.tab.7' is up to date
7c -FTVw -I../common ml.c
7c -FTVw -I../common common.c
dat.h:4 ml.c:2 7c: '../smtp/rfc822.tab.h' does not exist: ../smtp/rfc822.tab.h
mk: 7c -FTVw -I../common ml.c  : exit status=rc 12259: 7c 12261: error
dat.h:4 common.c:2 7c: '../smtp/rfc822.tab.h' does not exist: ../smtp/rfc822.tab.h
mk: for (i in ...  : exit status=rc 12253: mk 12255: error
mk: for (i in ...  : exit status=rc 12167: rc 12251: mk 12252: error
mk: for(i in 1a ...  : exit status=rc 9974: rc 12164: mk 12166: error
mk: test -e 7._cp ...  : exit status=rc 9442: mk 9957: error
mk: date for (i ...  : exit status=rc 7740: rc 9424: mk 9425: error

[-- Attachment #2: Type: message/rfc822, Size: 18326 bytes --]

[-- Attachment #2.1.1: Type: text/plain, Size: 297 bytes --]

Not sure if it was meant to be required after Amavect's latest mk
changes, but I had to do this after updating to the latest commit.  I
had followed FQA 5 to just do mk install under /sys/src, which wasn't
sufficient. I'll comment again if I need to do it down the road but
thought I'd mention it.

[-- Attachment #2.1.2: Type: message/rfc822, Size: 9445 bytes --]

From: ori@eigenstate.org
To: unobe@cpan.org, 9front@9front.org
Subject: Re: [9front] Cannot build system on latest rpi4 (libbio errors)
Date: Mon, 02 Nov 2020 12:55:12 -0800
Message-ID: <40159D135FE3FAB9F9029B243CC5BEF1@eigenstate.org>

> I ran sysupdate last night and attempted to rebuild the system, but it fails with some errors about libbio:
> 
> cpu% cd /sys/src
> cpu% mk install
> ...
> 7l  -o 7.out aliasmail.7 ../common/libcommon.a7
> ??none??: incompatible type signatures c0a3cfae(../common/libcommon.a7(sysnames_read)) and 9dd3ea4c(/arm64/lib/libbio.a(Bseek)) for Bseek
> badd: incompatible type signatures e8001435(../common/libcommon.a7(sysnames_read)) and 34728a43(/arm64/lib/libbio.a(Brdstr)) for Brdstr
> Binits: incompatible type signatures ac447a74(../common/libcommon.a7(sysnames_read)) and ce449e92(/arm64/lib/libbio.a(Bopen)) for Binit
> Bfdopen: incompatible type signatures 29f5223c(../common/libcommon.a7(sysnames_read)) and 63f5472a(/arm64/lib/libbio.a(Bopen)) for Bopen
> Bopen: incompatible type signatures bb9b2306(../common/libcommon.a7(sysnames_read)) and f9b3864(/arm64/lib/libbio.a(Bopen)) for Bterm
> ??none??: incompatible type signatures bb9b2306(../common/libcommon.a7(sysnames_read)) and f9b3864(/arm64/lib/libbio.a(Bfildes)) for Bfildes
> mk: 7l  -o ...  : exit status=rc 420270: 7l 420272: error
> mk: for (i in ...  : exit status=rc 420228: rc 420265: mk 420266: error
> mk: for(i in 1a ...  : exit status=rc 403330: rc 420225: mk 420227: error
> mk: test -e 7._cp ...  : exit status=rc 401470: mk 403313: error
> mk: date for (i ...  : exit status=rc 392586: rc 401452: mk 401453: error
> cpu%
> 
> I attempted to resolve by first rebuilding /sys/src/cc, then /sys/src/7c, as mentioned in FQA 5. I then tried 'mk clean', and 'mk libs', and still no success.  I would have thought that a 'mk clean' would have worked, but from the message above it looks like the build is using the system's libbio and it's expecting different type signatures. I looked at the commit log for 9front but didn't see anything recently that would modify libbio.
> 
> The only reference I found on google is here, with a different project: https://github.com/Orc/discount/issues/193

Let me guess, that's upas? I think the mkfiles there need
another look -- I don't think we rebuild libcommon when
we should.

Try 'cd /sys/src/cmd/upas && mk nuke'.

Re: [9front] Cannot build system on latest rpi4 (libbio errors)

[ori]

Quoth Romano <unobe@cpan.org>:
> To clarify, the failure wasn't exactly the same. Here is where it failed:
> 
> cd ../smtp && mk rfc822.tab.7
> mk: 'rfc822.tab.7' is up to date
> 7c -FTVw -I../common ml.c
> 7c -FTVw -I../common common.c
> dat.h:4 ml.c:2 7c: '../smtp/rfc822.tab.h' does not exist: ../smtp/rfc822.tab.h
> mk: 7c -FTVw -I../common ml.c  : exit status=rc 12259: 7c 12261: error
> dat.h:4 common.c:2 7c: '../smtp/rfc822.tab.h' does not exist: ../smtp/rfc822.tab.h
> mk: for (i in ...  : exit status=rc 12253: mk 12255: error
> mk: for (i in ...  : exit status=rc 12167: rc 12251: mk 12252: error
> mk: for(i in 1a ...  : exit status=rc 9974: rc 12164: mk 12166: error
> mk: test -e 7._cp ...  : exit status=rc 9442: mk 9957: error
> mk: date for (i ...  : exit status=rc 7740: rc 9424: mk 9425: error
> 

Hm.. can you try from a clean build?

Re: [9front] Cannot build system on latest rpi4 (libbio errors)

[Amavect]

On 11/21/2020 8:29 PM, ori@eigenstate.org wrote:
> Hm.. can you try from a clean build?

mk clean isn't cleaning *.tab.? as I had intended.
I'll look into this.

Thanks,
Amavect

Re: [9front] Cannot build system on latest rpi4 (libbio errors)

[Amavect]

[-- Attachment #1: Type: text/plain, Size: 1089 bytes --]

I've spotted 3 issues, two related and one unrelated.

First, ml/mkfile and send/mkfile have a recipe rebuilding only 
rfc822.tab.$O despite the targets being both ../smtp/rfc822.tab.h and 
../smtp/rfc822.tab.$O
If ../smtp/rfc822.tab.$O already exists, then ../smtp/rfc822.tab.h 
doesn't get rebuilt.
Second, the CLEANFILES=*.tab.? didn't actually cause *.tab.? to be 
deleted, despite mk's output clearly stating:
rm -f *.[05678qv] *.a[05678qv] y.tab.? lex.yy.c y.debug y.output 
[05678qv].??* smtpd smtp *.tab.?
I think this has to do with CLEANFILES being a variable, and the 
metacharacters were effectively quoted.
Since smtp/rfc822.tab.$O didn't get deleted, and the recipe only 
specifies 'mk smtp/rfc822.tab.$O', smtp/rfc822.tab.h wouldn't rebuild.
Neat interaction!
Fixing the rule and the CLEANFILES variable will work.

Unrelated, since I put <../mkupas at the top, the libcommon rule was 
being the default target. Instead, it should be the all rule. Moving the 
include after the mkmany or mkone include will work.

The attached patch should fix all of this.

Thanks,
Amavect

[-- Attachment #2: upas.diff --]
[-- Type: text/plain, Size: 7217 bytes --]

diff -r a7cc619b0017 sys/src/cmd/upas/Mail/mkfile
--- a/sys/src/cmd/upas/Mail/mkfile	Tue Nov 17 18:37:38 2020 -0800
+++ b/sys/src/cmd/upas/Mail/mkfile	Sat Nov 21 22:04:45 2020 -0600
@@ -1,5 +1,4 @@
 </$objtype/mkfile
-<../mkupas
 
 TARG=Mail
 OFILES=\
@@ -16,6 +15,7 @@
 BIN=$ABIN
 
 </sys/src/cmd/mkone
+<../mkupas
 
 $O.out: $OFILES
 	$LD -o $target  $LDFLAGS $OFILES
diff -r a7cc619b0017 sys/src/cmd/upas/alias/mkfile
--- a/sys/src/cmd/upas/alias/mkfile	Tue Nov 17 18:37:38 2020 -0800
+++ b/sys/src/cmd/upas/alias/mkfile	Sat Nov 21 22:04:45 2020 -0600
@@ -1,5 +1,4 @@
 </$objtype/mkfile
-<../mkupas
 
 TARG=aliasmail
 LIB=../common/libcommon.a$O
@@ -9,4 +8,5 @@
 	../common/sys.h\
 
 </sys/src/cmd/mkone
+<../mkupas
 CFLAGS=$CFLAGS -I../common
diff -r a7cc619b0017 sys/src/cmd/upas/bayes/mkfile
--- a/sys/src/cmd/upas/bayes/mkfile	Tue Nov 17 18:37:38 2020 -0800
+++ b/sys/src/cmd/upas/bayes/mkfile	Sat Nov 21 22:04:45 2020 -0600
@@ -1,9 +1,9 @@
 </$objtype/mkfile
-<../mkupas
 
 TARG=addhash bayes msgtok
 
 </sys/src/cmd/mkmany
+<../mkupas
 
 # msg tokenizer
 $O.regen: regcomp.$O dfa.$O
diff -r a7cc619b0017 sys/src/cmd/upas/binscripts/mkfile
--- a/sys/src/cmd/upas/binscripts/mkfile	Tue Nov 17 18:37:38 2020 -0800
+++ b/sys/src/cmd/upas/binscripts/mkfile	Sat Nov 21 22:04:45 2020 -0600
@@ -1,5 +1,4 @@
 </$objtype/mkfile
-<../mkupas
 
 RCFILES=\
 	isspam\
@@ -27,3 +26,5 @@
 
 $BIN/%: %.rc
 	cp $stem.rc $BIN/$stem
+
+<../mkupas
diff -r a7cc619b0017 sys/src/cmd/upas/filterkit/mkfile
--- a/sys/src/cmd/upas/filterkit/mkfile	Tue Nov 17 18:37:38 2020 -0800
+++ b/sys/src/cmd/upas/filterkit/mkfile	Sat Nov 21 22:04:45 2020 -0600
@@ -1,5 +1,4 @@
 </$objtype/mkfile
-<../mkupas
 
 TARG=\
 	deliver\
@@ -11,4 +10,5 @@
 OFILES=readaddrs.$O
 
 </sys/src/cmd/mkmany
+<../mkupas
 CFLAGS=$CFLAGS -I../common
diff -r a7cc619b0017 sys/src/cmd/upas/fs/mkfile
--- a/sys/src/cmd/upas/fs/mkfile	Tue Nov 17 18:37:38 2020 -0800
+++ b/sys/src/cmd/upas/fs/mkfile	Sat Nov 21 22:04:45 2020 -0600
@@ -1,5 +1,4 @@
 </$objtype/mkfile
-<../mkupas
 
 TARG=fs
 LIB=../common/libcommon.a$O
@@ -24,6 +23,7 @@
 	dat.h\
 
 </sys/src/cmd/mkone
+<../mkupas
 CFLAGS=$CFLAGS  -I../common
 
 acid:V:
diff -r a7cc619b0017 sys/src/cmd/upas/imap4d/mkfile
--- a/sys/src/cmd/upas/imap4d/mkfile	Tue Nov 17 18:37:38 2020 -0800
+++ b/sys/src/cmd/upas/imap4d/mkfile	Sat Nov 21 22:04:45 2020 -0600
@@ -1,5 +1,4 @@
 </$objtype/mkfile
-<../mkupas
 
 TARG=imap4d
 
@@ -31,3 +30,4 @@
 	fns.h\
 
 </sys/src/cmd/mkone
+<../mkupas
diff -r a7cc619b0017 sys/src/cmd/upas/marshal/mkfile
--- a/sys/src/cmd/upas/marshal/mkfile	Tue Nov 17 18:37:38 2020 -0800
+++ b/sys/src/cmd/upas/marshal/mkfile	Sat Nov 21 22:04:45 2020 -0600
@@ -1,5 +1,4 @@
 </$objtype/mkfile
-<../mkupas
 
 TARG=marshal
 LIB=../common/libcommon.a$O
@@ -7,5 +6,6 @@
 HFILES=../common/common.h
 
 </sys/src/cmd/mkone
+<../mkupas
 CFLAGS=$CFLAGS -I../common
 
diff -r a7cc619b0017 sys/src/cmd/upas/ml/mkfile
--- a/sys/src/cmd/upas/ml/mkfile	Tue Nov 17 18:37:38 2020 -0800
+++ b/sys/src/cmd/upas/ml/mkfile	Sat Nov 21 22:04:45 2020 -0600
@@ -1,5 +1,4 @@
 </$objtype/mkfile
-<../mkupas
 
 TARG=\
 	ml\
@@ -17,10 +16,11 @@
 	../smtp/rfc822.tab.h\
 
 </sys/src/cmd/mkmany
+<../mkupas
 CFLAGS=$CFLAGS -I../common
 
 $O.ml: ../smtp/rfc822.tab.$O
 $O.mlowner: ../smtp/rfc822.tab.$O
 
 ../smtp/rfc822.tab.h ../smtp/rfc822.tab.$O: ../smtp/rfc822.y
-	cd ../smtp && mk rfc822.tab.$O
+	cd ../smtp && mk rfc822.tab.h rfc822.tab.$O
diff -r a7cc619b0017 sys/src/cmd/upas/ned/mkfile
--- a/sys/src/cmd/upas/ned/mkfile	Tue Nov 17 18:37:38 2020 -0800
+++ b/sys/src/cmd/upas/ned/mkfile	Sat Nov 21 22:04:45 2020 -0600
@@ -1,5 +1,4 @@
 </$objtype/mkfile
-<../mkupas
 
 TARG=nedmail
 LIB=../common/libcommon.a$O
@@ -7,4 +6,5 @@
 HFILES=../common/common.h
 
 </sys/src/cmd/mkone
+<../mkupas
 CFLAGS=$CFLAGS -I../common
diff -r a7cc619b0017 sys/src/cmd/upas/pop3/mkfile
--- a/sys/src/cmd/upas/pop3/mkfile	Tue Nov 17 18:37:38 2020 -0800
+++ b/sys/src/cmd/upas/pop3/mkfile	Sat Nov 21 22:04:45 2020 -0600
@@ -1,9 +1,9 @@
 </$objtype/mkfile
-<../mkupas
 
 TARG=pop3
 LIB=../common/libcommon.a$O
 OFILES=pop3.$O
 
 </sys/src/cmd/mkone
+<../mkupas
 CFLAGS=$CFLAGS -I../common
diff -r a7cc619b0017 sys/src/cmd/upas/q/mkfile
--- a/sys/src/cmd/upas/q/mkfile	Tue Nov 17 18:37:38 2020 -0800
+++ b/sys/src/cmd/upas/q/mkfile	Sat Nov 21 22:04:45 2020 -0600
@@ -1,5 +1,4 @@
 </$objtype/mkfile
-<../mkupas
 
 TARG=\
 	qer\
@@ -12,4 +11,5 @@
 	../common/sys.h\
 
 </sys/src/cmd/mkmany
+<../mkupas
 CFLAGS=$CFLAGS -I../common
diff -r a7cc619b0017 sys/src/cmd/upas/qfrom/mkfile
--- a/sys/src/cmd/upas/qfrom/mkfile	Tue Nov 17 18:37:38 2020 -0800
+++ b/sys/src/cmd/upas/qfrom/mkfile	Sat Nov 21 22:04:45 2020 -0600
@@ -1,7 +1,7 @@
 </$objtype/mkfile
-<../mkupas
 
 TARG=qfrom
 OFILES=qfrom.$O
 
 </sys/src/cmd/mkone
+<../mkupas
diff -r a7cc619b0017 sys/src/cmd/upas/scanmail/mkfile
--- a/sys/src/cmd/upas/scanmail/mkfile	Tue Nov 17 18:37:38 2020 -0800
+++ b/sys/src/cmd/upas/scanmail/mkfile	Sat Nov 21 22:04:45 2020 -0600
@@ -1,5 +1,4 @@
 </$objtype/mkfile
-<../mkupas
 
 TARG=\
 	scanmail\
@@ -14,6 +13,7 @@
 	../common/sys.h\
 
 </sys/src/cmd/mkmany
+<../mkupas
 CFLAGS=$CFLAGS -I../common
 
 scanmail.$O: scanmail.c
diff -r a7cc619b0017 sys/src/cmd/upas/send/mkfile
--- a/sys/src/cmd/upas/send/mkfile	Tue Nov 17 18:37:38 2020 -0800
+++ b/sys/src/cmd/upas/send/mkfile	Sat Nov 21 22:04:45 2020 -0600
@@ -1,5 +1,4 @@
 </$objtype/mkfile
-<../mkupas
 
 TARG=\
 	send\
@@ -32,6 +31,7 @@
 	../common/sys.h\
 
 </sys/src/cmd/mkmany
+<../mkupas
 CFLAGS=$CFLAGS -I../common
 
 $O.send: $SOBJ $OFILES
@@ -42,4 +42,4 @@
 message.$O: ../smtp/rfc822.tab.h
 
 ../smtp/rfc822.tab.h ../smtp/rfc822.tab.$O: ../smtp/rfc822.y
-	cd ../smtp && mk rfc822.tab.$O
+	cd ../smtp && mk rfc822.tab.h rfc822.tab.$O
diff -r a7cc619b0017 sys/src/cmd/upas/smtp/mkfile
--- a/sys/src/cmd/upas/smtp/mkfile	Tue Nov 17 18:37:38 2020 -0800
+++ b/sys/src/cmd/upas/smtp/mkfile	Sat Nov 21 22:04:45 2020 -0600
@@ -1,5 +1,4 @@
 </$objtype/mkfile
-<../mkupas
 
 TARG=\
 	smtpd\
@@ -16,9 +15,10 @@
 
 TEST=parsetest
 
-CLEANFILES=*.tab.?
+CLEANFILES=rfc822.tab.h rfc822.tab.c smtpd.tab.c
 
 </sys/src/cmd/mkmany
+<../mkupas
 CFLAGS=$CFLAGS -I../common
 
 $O.smtpd:\
diff -r a7cc619b0017 sys/src/cmd/upas/spf/mkfile
--- a/sys/src/cmd/upas/spf/mkfile	Tue Nov 17 18:37:38 2020 -0800
+++ b/sys/src/cmd/upas/spf/mkfile	Sat Nov 21 22:04:45 2020 -0600
@@ -1,5 +1,4 @@
 </$objtype/mkfile
-<../mkupas
 
 TARG=spf
 
@@ -9,6 +8,7 @@
 	spf.$O\
 
 </sys/src/cmd/mkone
+<../mkupas
 
 mtest: dns.$O macro.$O mtest.$O
 	$LD $LDFLAGS -o $target $prereq
diff -r a7cc619b0017 sys/src/cmd/upas/unesc/mkfile
--- a/sys/src/cmd/upas/unesc/mkfile	Tue Nov 17 18:37:38 2020 -0800
+++ b/sys/src/cmd/upas/unesc/mkfile	Sat Nov 21 22:04:45 2020 -0600
@@ -1,7 +1,7 @@
 </$objtype/mkfile
-<../mkupas
 
 TARG=unesc
 OFILES=unesc.$O
 
 </sys/src/cmd/mkone
+<../mkupas
diff -r a7cc619b0017 sys/src/cmd/upas/vf/mkfile
--- a/sys/src/cmd/upas/vf/mkfile	Tue Nov 17 18:37:38 2020 -0800
+++ b/sys/src/cmd/upas/vf/mkfile	Sat Nov 21 22:04:45 2020 -0600
@@ -1,5 +1,4 @@
 </$objtype/mkfile
-<../mkupas
 
 TARG=vf
 LIB=../common/libcommon.a$O
@@ -9,4 +8,5 @@
 	../common/sys.h\
 
 </sys/src/cmd/mkone
+<../mkupas
 CFLAGS=$CFLAGS -I../common

Re: [9front] Cannot build system on latest rpi4 (libbio errors)

[Romano]

> On 11/21/2020 8:29 PM, ori@eigenstate.org wrote:
>> Hm.. can you try from a clean build?
> 
> mk clean isn't cleaning *.tab.? as I had intended.
> I'll look into this.

'mk clean', then 'mk install' works like a charm--no problems.
Apologies for the noise.

Re: [9front] Cannot build system on latest rpi4 (libbio errors)

[Romano]

[-- Attachment #1: Type: text/plain, Size: 297 bytes --]

Not sure if it was meant to be required after Amavect's latest mk
changes, but I had to do this after updating to the latest commit.  I
had followed FQA 5 to just do mk install under /sys/src, which wasn't
sufficient. I'll comment again if I need to do it down the road but
thought I'd mention it.

[-- Attachment #2: Type: message/rfc822, Size: 9445 bytes --]

From: ori@eigenstate.org
To: unobe@cpan.org, 9front@9front.org
Subject: Re: [9front] Cannot build system on latest rpi4 (libbio errors)
Date: Mon, 02 Nov 2020 12:55:12 -0800
Message-ID: <40159D135FE3FAB9F9029B243CC5BEF1@eigenstate.org>

> I ran sysupdate last night and attempted to rebuild the system, but it fails with some errors about libbio:
> 
> cpu% cd /sys/src
> cpu% mk install
> ...
> 7l  -o 7.out aliasmail.7 ../common/libcommon.a7
> ??none??: incompatible type signatures c0a3cfae(../common/libcommon.a7(sysnames_read)) and 9dd3ea4c(/arm64/lib/libbio.a(Bseek)) for Bseek
> badd: incompatible type signatures e8001435(../common/libcommon.a7(sysnames_read)) and 34728a43(/arm64/lib/libbio.a(Brdstr)) for Brdstr
> Binits: incompatible type signatures ac447a74(../common/libcommon.a7(sysnames_read)) and ce449e92(/arm64/lib/libbio.a(Bopen)) for Binit
> Bfdopen: incompatible type signatures 29f5223c(../common/libcommon.a7(sysnames_read)) and 63f5472a(/arm64/lib/libbio.a(Bopen)) for Bopen
> Bopen: incompatible type signatures bb9b2306(../common/libcommon.a7(sysnames_read)) and f9b3864(/arm64/lib/libbio.a(Bopen)) for Bterm
> ??none??: incompatible type signatures bb9b2306(../common/libcommon.a7(sysnames_read)) and f9b3864(/arm64/lib/libbio.a(Bfildes)) for Bfildes
> mk: 7l  -o ...  : exit status=rc 420270: 7l 420272: error
> mk: for (i in ...  : exit status=rc 420228: rc 420265: mk 420266: error
> mk: for(i in 1a ...  : exit status=rc 403330: rc 420225: mk 420227: error
> mk: test -e 7._cp ...  : exit status=rc 401470: mk 403313: error
> mk: date for (i ...  : exit status=rc 392586: rc 401452: mk 401453: error
> cpu%
> 
> I attempted to resolve by first rebuilding /sys/src/cc, then /sys/src/7c, as mentioned in FQA 5. I then tried 'mk clean', and 'mk libs', and still no success.  I would have thought that a 'mk clean' would have worked, but from the message above it looks like the build is using the system's libbio and it's expecting different type signatures. I looked at the commit log for 9front but didn't see anything recently that would modify libbio.
> 
> The only reference I found on google is here, with a different project: https://github.com/Orc/discount/issues/193

Let me guess, that's upas? I think the mkfiles there need
another look -- I don't think we rebuild libcommon when
we should.

Try 'cd /sys/src/cmd/upas && mk nuke'.

Re: [9front] Cannot build system on latest rpi4 (libbio errors)

[Romano]

>> I ran sysupdate last night and attempted to rebuild the system, but it fails with some errors about libbio:
>> ...
> 
> Let me guess, that's upas? I think the mkfiles there need
> another look -- I don't think we rebuild libcommon when
> we should.
> 
> Try 'cd /sys/src/cmd/upas && mk nuke'.

Ori, thank you! After doing that, I'm re-running 'mk install' from /sys/src and it looks like the system is continuing to build.

Re: [9front] Cannot build system on latest rpi4 (libbio errors)

[Romano]

[-- Attachment #1: Type: text/plain, Size: 155 bytes --]

Forgot to include the last (pertinent) paragraph:

Anyone have insight? Do I have to install the libraries first, then re-run 'mk install' under /sys/src?

[-- Attachment #2: Type: message/rfc822, Size: 10335 bytes --]

From: Romano <unobe@cpan.org>
To: 9front@9front.org
Subject: [9front] Cannot build system on latest rpi4 (libbio errors)
Date: Mon, 02 Nov 2020 12:49:55 -0800
Message-ID: <764142D9A571BCEFEEFF50F715733FB8@smtp.pobox.com>

I ran sysupdate last night and attempted to rebuild the system, but it fails with some errors about libbio:

cpu% cd /sys/src
cpu% mk install
...
7l  -o 7.out aliasmail.7 ../common/libcommon.a7
??none??: incompatible type signatures c0a3cfae(../common/libcommon.a7(sysnames_read)) and 9dd3ea4c(/arm64/lib/libbio.a(Bseek)) for Bseek
badd: incompatible type signatures e8001435(../common/libcommon.a7(sysnames_read)) and 34728a43(/arm64/lib/libbio.a(Brdstr)) for Brdstr
Binits: incompatible type signatures ac447a74(../common/libcommon.a7(sysnames_read)) and ce449e92(/arm64/lib/libbio.a(Bopen)) for Binit
Bfdopen: incompatible type signatures 29f5223c(../common/libcommon.a7(sysnames_read)) and 63f5472a(/arm64/lib/libbio.a(Bopen)) for Bopen
Bopen: incompatible type signatures bb9b2306(../common/libcommon.a7(sysnames_read)) and f9b3864(/arm64/lib/libbio.a(Bopen)) for Bterm
??none??: incompatible type signatures bb9b2306(../common/libcommon.a7(sysnames_read)) and f9b3864(/arm64/lib/libbio.a(Bfildes)) for Bfildes
mk: 7l  -o ...  : exit status=rc 420270: 7l 420272: error
mk: for (i in ...  : exit status=rc 420228: rc 420265: mk 420266: error
mk: for(i in 1a ...  : exit status=rc 403330: rc 420225: mk 420227: error
mk: test -e 7._cp ...  : exit status=rc 401470: mk 403313: error
mk: date for (i ...  : exit status=rc 392586: rc 401452: mk 401453: error
cpu%

I attempted to resolve by first rebuilding /sys/src/cc, then /sys/src/7c, as mentioned in FQA 5. I then tried 'mk clean', and 'mk libs', and still no success.  I would have thought that a 'mk clean' would have worked, but from the message above it looks like the build is using the system's libbio and it's expecting different type signatures. I looked at the commit log for 9front but didn't see anything recently that would modify libbio.

The only reference I found on google is here, with a different project: https://github.com/Orc/discount/issues/193

Re: [9front] Cannot build system on latest rpi4 (libbio errors)

[ori]

> I ran sysupdate last night and attempted to rebuild the system, but it fails with some errors about libbio:
> 
> cpu% cd /sys/src
> cpu% mk install
> ...
> 7l  -o 7.out aliasmail.7 ../common/libcommon.a7
> ??none??: incompatible type signatures c0a3cfae(../common/libcommon.a7(sysnames_read)) and 9dd3ea4c(/arm64/lib/libbio.a(Bseek)) for Bseek
> badd: incompatible type signatures e8001435(../common/libcommon.a7(sysnames_read)) and 34728a43(/arm64/lib/libbio.a(Brdstr)) for Brdstr
> Binits: incompatible type signatures ac447a74(../common/libcommon.a7(sysnames_read)) and ce449e92(/arm64/lib/libbio.a(Bopen)) for Binit
> Bfdopen: incompatible type signatures 29f5223c(../common/libcommon.a7(sysnames_read)) and 63f5472a(/arm64/lib/libbio.a(Bopen)) for Bopen
> Bopen: incompatible type signatures bb9b2306(../common/libcommon.a7(sysnames_read)) and f9b3864(/arm64/lib/libbio.a(Bopen)) for Bterm
> ??none??: incompatible type signatures bb9b2306(../common/libcommon.a7(sysnames_read)) and f9b3864(/arm64/lib/libbio.a(Bfildes)) for Bfildes
> mk: 7l  -o ...  : exit status=rc 420270: 7l 420272: error
> mk: for (i in ...  : exit status=rc 420228: rc 420265: mk 420266: error
> mk: for(i in 1a ...  : exit status=rc 403330: rc 420225: mk 420227: error
> mk: test -e 7._cp ...  : exit status=rc 401470: mk 403313: error
> mk: date for (i ...  : exit status=rc 392586: rc 401452: mk 401453: error
> cpu%
> 
> I attempted to resolve by first rebuilding /sys/src/cc, then /sys/src/7c, as mentioned in FQA 5. I then tried 'mk clean', and 'mk libs', and still no success.  I would have thought that a 'mk clean' would have worked, but from the message above it looks like the build is using the system's libbio and it's expecting different type signatures. I looked at the commit log for 9front but didn't see anything recently that would modify libbio.
> 
> The only reference I found on google is here, with a different project: https://github.com/Orc/discount/issues/193

Let me guess, that's upas? I think the mkfiles there need
another look -- I don't think we rebuild libcommon when
we should.

Try 'cd /sys/src/cmd/upas && mk nuke'.