I use 'info r' to read registers on mips64 target with dsp. something wrong. internal-error: void inf_ptrace_fetch_register(regcache*, int): Assertion `(size % sizeof (PTRACE_TYPE_RET)) == 0' failed. ----------------cut------------------------------- (gdb) b main Breakpoint 1 at 0x1200038b4 (gdb) r Starting program: /usr/bin/ls [Thread debugging using libthread_db enabled] Using host libthread_db library "/lib64/libthread_db.so.1". Breakpoint 1, 0x00000001200038b4 in main () (gdb) info r zero at v0 v1 R0 0000000000000000 0000000000000001 0000000000000000 000000fff7ffcf80 a0 a1 a2 a3 R4 0000000000000001 000000ffffffad98 000000ffffffada8 0000000000000000 a4 a5 a6 a7 R8 000000fff7f36820 000000fff7fdba18 000000ffffffad90 0000000000000001 t0 t1 t2 t3 R12 000000fff7d75140 000000000000000b 00000000f63d4e2e 000000fff7ffde28 s0 s1 s2 s3 R16 000000fff7f34cc8 000000012001d5e0 0000000000000000 0000000000000000 s4 s5 s6 s7 R20 000000012016d750 000000012015af48 0000000000000000 0000000000000000 t8 t9 k0 k1 R24 0000000000000000 00000001200038a0 0000000000000000 0000000000000000 gp sp s8 ra R28 00000001200438a0 000000ffffffa900 0000000000000000 000000fff7d931a0 status lo hi badvaddr ffffffff8500c8f3 0000000000000000 000000000000000a 000000fff7daefe8 cause pc 0000000010000024 00000001200038b4 fcsr fir hi1 lo1 00000000 00770501 0000000000000000 0000000000000000 hi2 lo2 hi3 lo3 0000000000000000 0000000000000000 0000000000000000 0000000000000000 dspctl restart ../../binutils-gdb_git_trunk/gdb/inf-ptrace.c:735: internal-error: void inf_ptrace_fetch_register(regcache*, int): Assertion `(size % sizeof (PTRACE_TYPE_RET)) == 0' failed. A problem internal to GDB has been detected, further debugging may prove unreliable. Quit this debugging session? (y or n) --------------------------------------------------------------------------
In mips64-dsp-linux.c:102 102 tdesc_create_reg (feature, "dspctl", 78, 1, NULL, 32, "int"); define the reg size is '32'。 but in inf-ptrace.c： 734 size = register_size (gdbarch, regnum); 735 gdb_assert ((size % sizeof (PTRACE_TYPE_RET)) == 0); the PTRACE_TYPE_RET is long, sizeof (PTRACE_TYPE_RET) is 8, size is 4, so triger the gdb_assert. we should define dspctl size is 64, like status/cause register. Changelog: 2017-10-12 Chenghua Xu <paul.hua.gm@gmail.com> * mips64-dsp-linux.c (initialize_tdesc_mips64_dsp_linux): Change dspctl from 32 to 64 bit. ----------------------------------------------------------------------- diff --git a/gdb/features/mips64-dsp-linux.c b/gdb/features/mips64-dsp-linux.c index 05317b7..b42a8ce 100644 --- a/gdb/features/mips64-dsp-linux.c +++ b/gdb/features/mips64-dsp-linux.c @@ -99,7 +99,7 @@ initialize_tdesc_mips64_dsp_linux (void) tdesc_create_reg (feature, "lo2", 75, 1, NULL, 64, "int"); tdesc_create_reg (feature, "hi3", 76, 1, NULL, 64, "int"); tdesc_create_reg (feature, "lo3", 77, 1, NULL, 64, "int"); - tdesc_create_reg (feature, "dspctl", 78, 1, NULL, 32, "int"); + tdesc_create_reg (feature, "dspctl", 78, 1, NULL, 64, "int"); feature = tdesc_create_feature (result, "org.gnu.gdb.mips.linux"); tdesc_create_reg (feature, "restart", 79, 1, "system", 64, "int");
CC: macro@linux-mips.org, cause git blame to you.
Hi Paul, Thank you for your bug report and the proposed fix. However the DSPControl register is architecturally 32-bit even on 64-bit hardware, like CP1 control registers (e.g. FCSR, etc.) and unlike CP0 registers in general. So I'd rather fix the problem the other way round. I'll look into it. NB I agree Status and Cause are inconsistent in that they are exposed as 64-bit while they are not. The reason is historical and boils down to the layout of the `g'/`G' RSP packets matching debug stubs which predate the introduction of XML register descriptions. This could be fixed in a way similar to how I adjusted the handling of FCSR and FIR not so long ago, however as a non-fatal issue this is not urgent. Maciej
Fix now posted: <https://patchwork.sourceware.org/patch/27300/>. NB as observed in the submission n64 support for the DSP state remains broken until <https://patchwork.kernel.org/patch/10402159/> has been applied to the Linux kernel and n32 support will remain broken due to a `ptrace' API limitation until both the kernel and GDB have adopted the regset API for DSP state access.
The master branch has been updated by Maciej W. Rozycki <macro@sourceware.org>: https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;h=1d7611244c140a1a08f3365cd5881120eba7749d commit 1d7611244c140a1a08f3365cd5881120eba7749d Author: Maciej W. Rozycki <macro@mips.com> Date: Wed May 16 20:43:30 2018 +0100 PR gdb/22286: linux-nat-trad: Support arbitrary register widths Update `fetch_register' and `store_register' code to support arbitrary register widths rather than only ones that are a multiply of the size of the `ptrace' data type used with PTRACE_PEEKUSR and PTRACE_POKEUSR requests to access registers. Remove associated assertions, correcting an issue with accessing the DSPControl (`$dspctl') register on n64 MIPS native targets: (gdb) print /x $dspctl .../gdb/linux-nat-trad.c:50: internal-error: void linux_nat_trad_target::fetch_register(regcache*, int): Assertion `(size % sizeof (PTRACE_TYPE_RET)) == 0' failed. A problem internal to GDB has been detected, further debugging may prove unreliable. Quit this debugging session? (y or n) n This is a bug, please report it. For instructions, see: <http://www.gnu.org/software/gdb/bugs/>. .../gdb/linux-nat-trad.c:50: internal-error: void linux_nat_trad_target::fetch_register(regcache*, int): Assertion `(size % sizeof (PTRACE_TYPE_RET)) == 0' failed. A problem internal to GDB has been detected, further debugging may prove unreliable. Create a core file of GDB? (y or n) n Command aborted. (gdb) All registers are now reported correctly and their architectural hardware widths respected: (gdb) print /x $dspctl $1 = 0x55aa33cc (gdb) info registers zero at v0 v1 R0 0000000000000000 0000000000000001 000000fff7ffeb20 0000000000000000 a0 a1 a2 a3 R4 0000000000000001 000000ffffffeaf8 000000ffffffeb08 0000000000000000 a4 a5 a6 a7 R8 000000fff7ee3800 000000fff7ede8f0 000000ffffffeaf0 2f2f2f2f2f2f2f2f t0 t1 t2 t3 R12 0000000000000437 0000000000000002 000000fff7ffd000 0000000120000ad0 s0 s1 s2 s3 R16 000000fff7ee2068 0000000120000e60 0000000000000000 0000000000000000 s4 s5 s6 s7 R20 0000000000521ec8 0000000000522608 0000000000000000 0000000000000000 t8 t9 k0 k1 R24 0000000000000000 0000000120000d9c 0000000000000000 0000000000000000 gp sp s8 ra R28 0000000120019030 000000ffffffe990 000000ffffffe990 000000fff7d5b88c status lo hi badvaddr 0000000000109cf3 0000000000005ea5 0000000000000211 000000fff7fc6fe0 cause pc 0000000000800024 0000000120000dbc fcsr fir hi1 lo1 00000000 00f30000 0000000000000000 0101010101010101 hi2 lo2 hi3 lo3 0202020202020202 0303030303030303 0404040404040404 0505050505050505 dspctl restart 55aa33cc 0000000000000000 (gdb) NB due to the lack of access to 64-bit DSP hardware all DSP register values in the dumps are artificial and have been created with a debug change applied to the kernel handler of the `ptrace' syscall. The use of `store_unsigned_integer' and `extract_unsigned_integer' unconditionally in all cases rather than when actual data occupies a part of the data quantity exchanged with `ptrace' makes code perhaps marginally slower, however I think avoiding it is not worth code obfuscation it would cause. If this turns out unfounded, then there should be no problem with optimizing this code later. gdb/ PR gdb/22286 * linux-nat-trad.c (linux_nat_trad_target::fetch_register): Also handle registers whose width is not a multiple of PTRACE_TYPE_RET. (linux_nat_trad_target::store_register): Likewise.
Fix now committed, closing bug.