Skip to contentSkip to conversation

Fix pahole to emit kernel module BTF variables

PATCH Applied Andrii Nakryiko 5 participants 8 messages Dec 11, 2020 – Dec 15, 2020 BPF

lore
Share & tools
lore

Andrii Nakryiko

author maintainer
#0 lore raw

cover Fix pahole to emit kernel module BTF variables

Two bug fixes to make pahole emit correct kernel module BTF variable information.

Cc Hao Luo <[email protected]>
Cc Jiri Olsa <[email protected]>

Andrii Nakryiko (2):
btf_encoder: fix BTF variable generation for kernel modules
btf_encoder: fix skipping per-CPU variables at offset 0

 btf_encoder.c | 61 +++++++++++++++++++++++++++++++++------------------
 libbtf.c      |  1 +
 libbtf.h      |  1 +
 3 files changed, 42 insertions(+), 21 deletions(-)
Signature · 1 line
2.24.1
author maintainer
#1 lore raw

PATCH 1/2 btf_encoder: fix BTF variable generation for kernel modules

Fix pahole's logic for determining per-CPU variables. For vmlinux, btfe->percpu_base_addr is always 0, so it didn't matter at which point to subtract it to get offset that later was matched against corresponding ELF symbol.

For kernel module, though, the situation is different. Kernel module's per-CPU data section has non-zero offset, which is taken into account in all DWARF variable addresses calculation. For such cases, it's important to subtract section offset (btfe->percpu_base_addr) before ELF symbol look up is performed.

This patch also records per-CPU data section size and uses it for early filtering of non-per-CPU variables by their address.

Fixes 2e719cca6672 ("btf_encoder: revamp how per-CPU variables are encoded")
---
 btf_encoder.c | 21 ++++++++++++++++-----
 libbtf.c      |  1 +
 libbtf.h      |  1 +
 3 files changed, 18 insertions(+), 5 deletions(-)
btf_encoder.c +16 -5
index c40f059580da..a7d484765ce2 100644
@@ -651,7 +651,7 @@ int cu__encode_btf(struct cu *cu, int verbose, bool force,
 		printf("search cu '%s' for percpu global variables.\n", cu->name);
 
 	cu__for_each_variable(cu, core_id, pos) {
-		uint32_t size, type, linkage, offset;
+		uint32_t size, type, linkage;
 		const char *name;
 		uint64_t addr;
 		int id;
@@ -665,12 +665,24 @@ int cu__encode_btf(struct cu *cu, int verbose, bool force,
 
 		/* addr has to be recorded before we follow spec */
 		addr = var->ip.addr;
-		if (var->spec)
-			var = var->spec;
+
+		/* DWARF takes into account .data..percpu section offset
+		 * within its segment, which for vmlinux is 0, but for kernel
+		 * modules is >0. ELF symbols, on the other hand, don't take
+		 * into account these offsets (as they are relative to the
+		 * section start), so to match DWARF and ELF symbols we need
+		 * to negate the section base address here.
+		 */
+		if (addr < btfe->percpu_base_addr || addr >= btfe->percpu_base_addr + btfe->percpu_sec_sz)
+			continue;
+		addr -= btfe->percpu_base_addr;
 
 		if (!percpu_var_exists(addr, &size, &name))
 			continue; /* not a per-CPU variable */
 
+		if (var->spec)
+			var = var->spec;
+
 		if (var->ip.tag.type == 0) {
 			fprintf(stderr, "error: found variable '%s' in CU '%s' that has void type\n",
 				name, cu->name);
@@ -701,8 +713,7 @@ int cu__encode_btf(struct cu *cu, int verbose, bool force,
 		 * add a BTF_VAR_SECINFO in btfe->percpu_secinfo, which will be added into
 		 * btfe->types later when we add BTF_VAR_DATASEC.
 		 */
-		offset = addr - btfe->percpu_base_addr;
-		id = btf_elf__add_var_secinfo(&btfe->percpu_secinfo, id, offset, size);
+		id = btf_elf__add_var_secinfo(&btfe->percpu_secinfo, id, addr, size);
 		if (id < 0) {
 			err = -1;
 			fprintf(stderr, "error: failed to encode section info for variable '%s' at addr 0x%lx\n",
libbtf.c +1 -0
index 246762c4b4e1..16e1d451e433 100644
@@ -170,6 +170,7 @@ try_as_raw_btf:
 	}
 	btfe->percpu_shndx = elf_ndxscn(sec);
 	btfe->percpu_base_addr = shdr.sh_addr;
+	btfe->percpu_sec_sz = shdr.sh_size;
 
 	return btfe;
 
libbtf.h +1 -0
index 71f6cecbea93..191f5862a695 100644
@@ -26,6 +26,7 @@ struct btf_elf {
 	bool		  raw_btf; // "/sys/kernel/btf/vmlinux"
 	uint32_t	  percpu_shndx;
 	uint64_t	  percpu_base_addr;
+	uint64_t	  percpu_sec_sz;
 	struct btf	  *btf;
 	struct btf	  *base_btf;
 };
Signature · 1 line
2.24.1
author maintainer
#2 lore raw

PATCH 2/2 btf_encoder: fix skipping per-CPU variables at offset 0

In reply to Andrii Nakryiko

Adjust pahole logic of skipping any per-CPU symbol with offset 0, which is especially bad for kernel modules, because it most certainly skips the very first per-CPU variable.

Instead, do collect per-CPU ELF symbol with 0 offset, but do extra check for non-kernel module case by verifying that ELF symbol name and DWARF variable name match. Due to the bug of DWARF name of variable sometimes being NULL, this is necessarily too pessimistic check (e.g., on my vmlinux image, fixed_percpu_data variable is still not emitted due to missing DWARF variable name), it allows to emit data for all module per-CPU variables.

Fixes f3d9054ba8ff ("btf_encoder: Teach pahole to store percpu variables in vmlinux BTF.")
---
 btf_encoder.c | 40 ++++++++++++++++++++++++----------------
 1 file changed, 24 insertions(+), 16 deletions(-)
btf_encoder.c +24 -16
index a7d484765ce2..1d7817078f89 100644
@@ -412,21 +412,6 @@ static int collect_percpu_var(struct btf_elf *btfe, GElf_Sym *sym)
 		return 0;
 
 	addr = elf_sym__value(sym);
-	/*
-	 * Store only those symbols that have allocated space in the percpu section.
-	 * This excludes the following three types of symbols:
-	 *
-	 *  1. __ADDRESSABLE(sym), which are forcely emitted as symbols.
-	 *  2. __UNIQUE_ID(prefix), which are introduced to generate unique ids.
-	 *  3. __exitcall(fn), functions which are labeled as exit calls.
-	 *
-	 * In addition, the variables defined using DEFINE_PERCPU_FIRST are
-	 * also not included, which currently includes:
-	 *
-	 *  1. fixed_percpu_data
-	 */
-	if (!addr)
-		return 0;
 
 	size = elf_sym__size(sym);
 	if (!size)
@@ -652,7 +637,7 @@ int cu__encode_btf(struct cu *cu, int verbose, bool force,
 
 	cu__for_each_variable(cu, core_id, pos) {
 		uint32_t size, type, linkage;
-		const char *name;
+		const char *name, *dwarf_name;
 		uint64_t addr;
 		int id;
 
@@ -680,6 +665,29 @@ int cu__encode_btf(struct cu *cu, int verbose, bool force,
 		if (!percpu_var_exists(addr, &size, &name))
 			continue; /* not a per-CPU variable */
 
+		/* A lot of "special" DWARF variables (e.g, __UNIQUE_ID___xxx)
+		 * have addr == 0, which is the same as, say, valid
+		 * fixed_percpu_data per-CPU variable. To distinguish between
+		 * them, additionally compare DWARF and ELF symbol names. If
+		 * DWARF doesn't provide proper name, pessimistically assume
+		 * bad variable.
+		 *
+		 * Examples of such special variables are:
+		 *
+		 *  1. __ADDRESSABLE(sym), which are forcely emitted as symbols.
+		 *  2. __UNIQUE_ID(prefix), which are introduced to generate unique ids.
+		 *  3. __exitcall(fn), functions which are labeled as exit calls.
+		 *
+		 *  This is relevant only for vmlinux image, as for kernel
+		 *  modules per-CPU data section has non-zero offset so all
+		 *  per-CPU symbols have non-zero values.
+		 */
+		if (var->ip.addr == 0) {
+			dwarf_name = variable__name(var, cu);
+			if (!dwarf_name || strcmp(dwarf_name, name))
+				continue;
+		}
+
 		if (var->spec)
 			var = var->spec;
 
Signature · 1 line
2.24.1
#3 lore raw

Re: [PATCH dwarves 2/2] btf_encoder: fix skipping per-CPU variables at offset 0

Andrii, sorry for the late reply. This change looks good. It makes much more sense than before.

Thanks,
Hao

Quoted 86 lines · Andrii Nakryiko

Adjust pahole logic of skipping any per-CPU symbol with offset 0, which is especially bad for kernel modules, because it most certainly skips the very first per-CPU variable.

Instead, do collect per-CPU ELF symbol with 0 offset, but do extra check for non-kernel module case by verifying that ELF symbol name and DWARF variable name match. Due to the bug of DWARF name of variable sometimes being NULL, this is necessarily too pessimistic check (e.g., on my vmlinux image, fixed_percpu_data variable is still not emitted due to missing DWARF variable name), it allows to emit data for all module per-CPU variables.

Fixes f3d9054ba8ff ("btf_encoder: Teach pahole to store percpu variables in vmlinux BTF.")
---
 btf_encoder.c | 40 ++++++++++++++++++++++++----------------
 1 file changed, 24 insertions(+), 16 deletions(-)
btf_encoder.c +24 -17
index a7d484765ce2..1d7817078f89 100644
@@ -412,21 +412,6 @@ static int collect_percpu_var(struct btf_elf *btfe, GElf_Sym *sym)
                return 0;
 
        addr = elf_sym__value(sym);
-       /*
-        * Store only those symbols that have allocated space in the percpu section.
-        * This excludes the following three types of symbols:
-        *
-        *  1. __ADDRESSABLE(sym), which are forcely emitted as symbols.
-        *  2. __UNIQUE_ID(prefix), which are introduced to generate unique ids.
-        *  3. __exitcall(fn), functions which are labeled as exit calls.
-        *
-        * In addition, the variables defined using DEFINE_PERCPU_FIRST are
-        * also not included, which currently includes:
-        *
-        *  1. fixed_percpu_data
-        */
-       if (!addr)
-               return 0;
 
        size = elf_sym__size(sym);
        if (!size)
@@ -652,7 +637,7 @@ int cu__encode_btf(struct cu *cu, int verbose, bool force,
 
        cu__for_each_variable(cu, core_id, pos) {
                uint32_t size, type, linkage;
-               const char *name;
+               const char *name, *dwarf_name;
                uint64_t addr;
                int id;
 
@@ -680,6 +665,29 @@ int cu__encode_btf(struct cu *cu, int verbose, bool force,
                if (!percpu_var_exists(addr, &size, &name))
                        continue; /* not a per-CPU variable */
 
+               /* A lot of "special" DWARF variables (e.g, __UNIQUE_ID___xxx)
+                * have addr == 0, which is the same as, say, valid
+                * fixed_percpu_data per-CPU variable. To distinguish between
+                * them, additionally compare DWARF and ELF symbol names. If
+                * DWARF doesn't provide proper name, pessimistically assume
+                * bad variable.
+                *
+                * Examples of such special variables are:
+                *
+                *  1. __ADDRESSABLE(sym), which are forcely emitted as symbols.
+                *  2. __UNIQUE_ID(prefix), which are introduced to generate unique ids.
+                *  3. __exitcall(fn), functions which are labeled as exit calls.
+                *
+                *  This is relevant only for vmlinux image, as for kernel
+                *  modules per-CPU data section has non-zero offset so all
+                *  per-CPU symbols have non-zero values.
+                */
+               if (var->ip.addr == 0) {
+                       dwarf_name = variable__name(var, cu);
+                       if (!dwarf_name || strcmp(dwarf_name, name))
+                               continue;
+               }
+
                if (var->spec)
                        var = var->spec;
 
--

2.24.1

Acked-by on the v1 series
#4 lore raw
In reply to Andrii Nakryiko
Quoted 9 lines · Andrii Nakryiko

Two bug fixes to make pahole emit correct kernel module BTF variable information.

Cc Hao Luo <[email protected]>
Cc Jiri Olsa <[email protected]>

Andrii Nakryiko (2):
btf_encoder: fix BTF variable generation for kernel modules
btf_encoder: fix skipping per-CPU variables at offset 0

jirka

Quoted 9 lines
 btf_encoder.c | 61 +++++++++++++++++++++++++++++++++------------------
 libbtf.c      |  1 +
 libbtf.h      |  1 +
 3 files changed, 42 insertions(+), 21 deletions(-)
Signature · 1 line
2.24.1
#5 lore raw

Em Sun, Dec 13, 2020 at 09:27:57PM +0100, Jiri Olsa escreveu:

Quoted 12 lines · Andrii Nakryiko
Quoted 9 lines

Two bug fixes to make pahole emit correct kernel module BTF variable information.

Cc Hao Luo <[email protected]>
Cc Jiri Olsa <[email protected]>

Andrii Nakryiko (2):
btf_encoder: fix BTF variable generation for kernel modules
btf_encoder: fix skipping per-CPU variables at offset 0

Thanks, applied.

- Arnaldo

#6 lore raw
Quoted 9 lines · Arnaldo Carvalho de Melo

Em Sun, Dec 13, 2020 at 09:27:57PM +0100, Jiri Olsa escreveu:

Quoted 7 lines · Andrii Nakryiko
Quoted 6 lines

Two bug fixes to make pahole emit correct kernel module BTF variable information.

Cc Hao Luo <[email protected]>
Cc Jiri Olsa <[email protected]>
Quoted 8 lines
Quoted 5 lines
Quoted 3 lines

Andrii Nakryiko (2):
btf_encoder: fix BTF variable generation for kernel modules
btf_encoder: fix skipping per-CPU variables at offset 0

Thanks, applied.

Thanks, Arnaldo. I'm not seeing them on gitolite.kernel.org/pub/scm/devel/pahole/pahole.git, did you push them somewhere else?

Quoted 2 lines

- Arnaldo

#7 lore raw

Em Mon, Dec 14, 2020 at 05:28:22PM -0800, Andrii Nakryiko escreveu:

Quoted 15 lines
Quoted 13 lines · Arnaldo Carvalho de Melo

Em Sun, Dec 13, 2020 at 09:27:57PM +0100, Jiri Olsa escreveu:

Quoted 11 lines · Andrii Nakryiko
Quoted 9 lines

Two bug fixes to make pahole emit correct kernel module BTF variable information.

Cc Hao Luo <[email protected]>
Cc Jiri Olsa <[email protected]>

Andrii Nakryiko (2):
btf_encoder: fix BTF variable generation for kernel modules
btf_encoder: fix skipping per-CPU variables at offset 0

Quoted 8 lines
Quoted 4 lines
Quoted 1 line

Thanks, applied.

Thanks, Arnaldo. I'm not seeing them on gitolite.kernel.org/pub/scm/devel/pahole/pahole.git, did you push them somewhere else?

I need to push it out, I'm doing kernel builds and tests and will incorporate Hao's Acked-by.

- Arnaldo