re: mac68k kern-INSTALL vs GCC7?
"John D. Baker" writes: > On Fri, 15 Feb 2019, John D. Baker wrote: > > > On Fri, 15 Feb 2019, Jaromir Dolecek wrote: > > > > > Maybe something like this? > > > > > > https://www.netbsd.org/~jdolecek/mac68k_intr_gcc7.diff > > > > With this the build succeeds. > > > > Will check Matthew Green's version after the next round of updates. > > This also allows -current mac68k build to complete with -V HAVE_GCC=7. > > (Don't know why his email addressed to me and Cc:ed to current-users > hasn't reached me--only saw it in current-users.) yeah - this is the only m68k build breaking change i didn't commit already because i was wanting a tester. all the other m68k ports should build and maybe work. amiga had reasonable atf results i believe. .mrg.
Re: mac68k kern-INSTALL vs GCC7?
On Fri, 15 Feb 2019, John D. Baker wrote: > On Fri, 15 Feb 2019, Jaromir Dolecek wrote: > > > Maybe something like this? > > > > https://www.netbsd.org/~jdolecek/mac68k_intr_gcc7.diff > > With this the build succeeds. > > Will check Matthew Green's version after the next round of updates. This also allows -current mac68k build to complete with -V HAVE_GCC=7. (Don't know why his email addressed to me and Cc:ed to current-users hasn't reached me--only saw it in current-users.) -- |/"\ John D. Baker, KN5UKS NetBSD Darwin/MacOS X |\ / jdbaker[snail]consolidated[flyspeck]net OpenBSDFreeBSD | X No HTML/proprietary data in email. BSD just sits there and works! |/ \ GPGkeyID: D703 4A7E 479F 63F8 D3F4 BD99 9572 8F23 E4AD 1645
Re: mac68k kern-INSTALL vs GCC7?
In article <4559.1550263...@splode.eterna.com.au>, matthew green wrote: >this is my preferred change vs jarmoir's. please test it. > > https://www.netbsd.org/~mrg/mac68k-intr.diff > >i was also planning on adding an assert that MAX_INAME_LENGTH is >less than (eintrnames - intrnames). I would use sizeof() in the memcpy. christos
Re: mac68k kern-INSTALL vs GCC7?
On Fri, 15 Feb 2019, Jaromir Dolecek wrote: > Maybe something like this? > > https://www.netbsd.org/~jdolecek/mac68k_intr_gcc7.diff With this the build succeeds. Will check Matthew Green's version after the next round of updates. -- |/"\ John D. Baker, KN5UKS NetBSD Darwin/MacOS X |\ / jdbaker[snail]consolidated[flyspeck]net OpenBSDFreeBSD | X No HTML/proprietary data in email. BSD just sits there and works! |/ \ GPGkeyID: D703 4A7E 479F 63F8 D3F4 BD99 9572 8F23 E4AD 1645
re: mac68k kern-INSTALL vs GCC7?
this is my preferred change vs jarmoir's. please test it. https://www.netbsd.org/~mrg/mac68k-intr.diff i was also planning on adding an assert that MAX_INAME_LENGTH is less than (eintrnames - intrnames). .mrg.
Re: mac68k kern-INSTALL vs GCC7?
Le ven. 15 févr. 2019 à 17:33, John D. Baker a écrit : > > Building for mac68k with -V HAVE_GCC=7 produces the following error: > > /x/current/src/sys/arch/mac68k/mac68k/intr.c:135:2: note: in expansion of > macro 'memcpy' > memcpy(g_inames, inames, MAX_INAME_LENGTH); > ^~ > cc1: all warnings being treated as errors > *** [intr.o] Error code 1 > nbmake[2]: stopped in > /r0/build/current/obj/mac68k/sys/arch/mac68k/compile/INSTALL > 1 error > nbmake[2]: stopped in > /r0/build/current/obj/mac68k/sys/arch/mac68k/compile/INSTALL > [...] The code there boilds down to memcpy(, "somestring", 53); which while actually being (upon inspection) fine, it is very difficult to parse and compiler warns correctly. Maybe something like this? https://www.netbsd.org/~jdolecek/mac68k_intr_gcc7.diff Jaromir
mac68k kern-INSTALL vs GCC7?
Building for mac68k with -V HAVE_GCC=7 produces the following error: [...] --- kern-INSTALL --- In file included from /x/current/src/sys/sys/systm.h:455:0, from /x/current/src/sys/sys/timevar.h:66, from /x/current/src/sys/sys/time.h:307, from /x/current/src/sys/sys/param.h:145, from /x/current/src/sys/arch/mac68k/mac68k/intr.c:39: /x/current/src/sys/arch/mac68k/mac68k/intr.c: In function 'intr_init': /x/current/src/sys/lib/libkern/libkern.h:380:26: error: '__builtin_memcpy' writing 53 bytes into a region of size 4 overflows the destination [-Werror=stringop-overflow=] #define memcpy(d, s, l) __builtin_memcpy(d, s, l) ^ /x/current/src/sys/arch/mac68k/mac68k/intr.c:135:2: note: in expansion of macro 'memcpy' memcpy(g_inames, inames, MAX_INAME_LENGTH); ^~ cc1: all warnings being treated as errors *** [intr.o] Error code 1 nbmake[2]: stopped in /r0/build/current/obj/mac68k/sys/arch/mac68k/compile/INSTALL 1 error nbmake[2]: stopped in /r0/build/current/obj/mac68k/sys/arch/mac68k/compile/INSTALL [...] Other m68k arches I routinely build (mvme68k, amiga) build OK with GCC7. No runtime testing until I can excavate the machines as well as space where they can be set up to run... -- |/"\ John D. Baker, KN5UKS NetBSD Darwin/MacOS X |\ / jdbaker[snail]consolidated[flyspeck]net OpenBSDFreeBSD | X No HTML/proprietary data in email. BSD just sits there and works! |/ \ GPGkeyID: D703 4A7E 479F 63F8 D3F4 BD99 9572 8F23 E4AD 1645